CVE-2026-47865 — VMware Avi Load Balancer: unauthenticated control-plane authentication bypass (CVSS 9.8), no workaround

Broadcom's VMSA-2026-0005 (2026-07-14, last updated 2026-07-15) patches seven vulnerabilities in VMware Avi Load Balancer — the load-balancing/application-delivery product formerly sold as NSX Advanced Load Balancer and widely deployed in enterprise and government data-centre fabric across Europe. The load-bearing flaw, CVE-2026-47865 (CVSS 9.8), is an authentication bypass on the Avi Controller: "a malicious user with network access may be able to access the Avi Control plane by bypassing the authentication mechanism" — no credentials, no user interaction (Broadcom PSIRT, 2026-07-14). The advisory ships six companion flaws that require a prior foothold: two high-privilege remote code-execution bugs (CVE-2026-47867, CVE-2026-47869, both CVSS 8.7, PR:H), a low-privilege authenticated directory traversal (CVE-2026-47871, CVSS 8.8), an authorization bypass (CVE-2026-47866, CVSS 8.3), a local-to-root privilege escalation (CVE-2026-47868, CVSS 7.8) and a further privilege escalation (CVE-2026-47870, CVSS 7.1). Broadcom states no workarounds exist (Broadcom PSIRT, 2026-07-14); the German trade press summarised it as attackers being able to bypass authentication and authorization (heise Security, 2026-07-17).

No in-the-wild exploitation has been reported, but two facts raise this above the routine patch cycle: the reporter is the NATO NCSC (a direct constituency-provenance signal), and there is no mitigation short of upgrading. Because the Avi Controller is the management and orchestration plane for the load-balancing fabric, an unauthenticated bypass there is a direct path to reconfiguring traffic routing and TLS termination for every service behind the load balancer — an interception and traffic-manipulation position, not merely appliance compromise.

SonicWall SMA 1000 zero-day exploitation (CVE-2026-15409/-15410): Volexity reconstructs UTA0533's full appliance-to-network kill chain

UPDATE · originally covered CVE-2026-15409 — SonicWall SMA1000: unauthenticated SSRF (CVSS 10.0) chained to post-auth code injection, actively exploited (2026-07-14)

The original entry recorded SonicWall's confirmation that CVE-2026-15409/-15410 were being exploited as zero-days and directed emergency patching. Volexity has now published the reconstructed intrusion, attributed it to an actor it tracks as UTA0533, and shown that patching alone is insufficient — the delta below is the full kill chain, the on-appliance implants, and the compromise-response guidance the terse advisory did not carry (Volexity, 2026-07-17).

Volexity was engaged after suspect authentication and lateral movement were seen originating from SonicWall SMA 1000 appliances (models 6210/7210/8200v); the earliest sign of compromise was 2026-06-22, weeks before SonicWall's 2026-07-14 disclosure (Volexity, 2026-07-17). SonicWall's PSIRT confirms it "has investigated multiple cases indicating the active exploitation of the vulnerabilities described in this advisory" (SonicWall SNWLID-2026-0008, 2026-07-14), and Rapid7's MDR team independently found the same two zero-days under attack (Rapid7, 2026-07-16).

Initial access (T1190, T1133). CVE-2026-15409 is a pre-authentication server-side request forgery in the SMA 1000 /wsproxy endpoint. A crafted WebSocket-upgrade request establishes a tunnel from the unauthenticated external attacker straight to services that are supposed to be reachable only on the appliance's own loopback — Volexity confirms "no valid SMA session cookie was required during this process" (Volexity, 2026-07-17). Through the tunnel the actor reached the appliance's bundled CouchDB/Erlang services and a localhost-only control service; the shipped CouchDB carries hardcoded admin:admin credentials, and the control service's authentication password is derivable from a device UUID, so the SSRF turns both into part of the external attack surface.

Privilege escalation (T1068). CVE-2026-15410 is a path traversal in the hotfix-rollback workflow: the sysCtrl.execRemoveHotfix operation builds a rollback path from caller-controlled input and hands it to /usr/local/bin/remove_hotfix, which then executes it. A rollback name containing directory-traversal sequences resolves outside the intended rollback directory and runs an attacker-staged script as root.

Persistence and implants (T1055, T1505.003, T1090.003, T1037.004). With root, UTA0533 dropped a setuid helper and a Python loader Volexity calls KNUCKLEBALL, which injects two JAR archives into the appliance's legitimate workplace process: the open-source Suo5 HTTP proxy-forwarder and a Behinder-like Java webshell Volexity calls ORANGETAIL. Persistence was established by adding a call to the loader inside the appliance's workplace init script, and the NGINX Unit configuration was rewritten to add routes that proxy attacker-chosen (arbitrary) request paths to the injected webshell and proxy — so hunting for a fixed URL is the wrong shape; the behaviour is unexpected route entries in the appliance's own reverse-proxy configuration.

Credential access and lateral movement (T1040, T1059). The actor ran tcpdump from a script staged in the appliance's temp directory to capture unencrypted LDAP traffic (TCP 389), harvesting directory credentials off the wire (Volexity, 2026-07-17). Rapid7's engagement observed the actor then "quickly shifted to lateral movement, pivoting from the compromised appliance directly into the internal corporate network" (Rapid7, 2026-07-16). How far that onward movement reached differs across the two IR firms' cases: Volexity concludes that in the appliances it investigated, "available evidence suggests the threat actor was less successful moving laterally or gaining access to other systems" (Volexity, 2026-07-17) — so treat a foothold on the appliance as a demonstrated launch point for internal movement, but not evidence that deep lateral movement always succeeds.

Garante fines Wind Tre EUR 1.7M over a vishing-enabled API-enumeration breach that exposed 365,048 telco customers

The Garante's decision gives a rare, fully technical account of a telco breach. Initial access was voice social engineering: attackers phoned staff at two retail points of sale, posed as internal support technicians, and "convinced operators at two retail points of sale to allow access to company systems" (Garante, 2026-07-16). That remote access yielded the point-of-sale device's installed client digital certificate plus login credentials — reportedly recoverable in cleartext from the desktop or browser rather than held in an OS certificate store — which the attackers then used as valid, MFA-satisfied access to a customer-facing web application. In the first incident that access ran 66 targeted lookups (~23 customers). In the second, days later, the attackers pivoted from the primary (protected) search API to an unprotected secondary API invoked by the same search function and "executed about 2 million total requests following an enumeration logic, i.e. progressively incrementing the customer code identifier ('customerId')," compromising 365,048 customers and, for 41,359 of them, payment-instrument data (Garante, 2026-07-16). The Garante rejected Wind Tre's defense that its API design followed OWASP practice, finding the enumeration-reachable secondary endpoints were "reasonably identifiable" by a vulnerability assessment and penetration test scoped to the API surface — not just the primary documented interfaces.

Basel utility IWB: ~40,000 customer records exfiltrated in a breach of a third-party service provider

Industrielle Werke Basel (IWB) — the canton-owned Basel multi-utility supplying electricity, gas, water, district heating and telecom/fibre — disclosed on 15 July 2026 that an external service provider it uses was compromised and roughly 40,000 customer records were exfiltrated from the provider's environment (Netzwoche, 2026-07-15). The stolen data comprises customer names and addresses plus technical smart-meter attributes (meter serial numbers and installation characteristics); IWB states that email addresses, phone numbers, energy-consumption data and billing/payment data were not part of the exposure, so no consumption-pattern inference is possible from what was taken (SwissCybersecurity.net, 2026-07-15). IWB's own IT and OT/grid systems were unaffected and energy/water supply continuity was not disrupted — the compromise is scoped to the provider's systems and the customer-data feed IWB shares with it (Netzwoche, 2026-07-15). The provider detected and notified IWB, which audited access, reviewed logs and pre-emptively restricted its data exchange with the affected provider; the Basel-Stadt cantonal data protection officer assessed the misuse risk as low (Watson.ch, 2026-07-15). No provider name, threat-actor claim or initial-access vector has been disclosed, and no matching leak-site listing was found for Switzerland in-window.

Proofpoint: OAuth client ID spoofing validates stolen Entra ID credentials at scale without writing a successful sign-in log

Proofpoint's Threat Research team documented a stealthy authentication-evasion technique — OAuth client ID spoofing — being independently weaponised by two distinct clusters against Microsoft Entra ID (Proofpoint, 2026-07-13). The mechanism abuses the legacy Resource Owner Password Credentials (ROPC) flow: an attacker POSTs a username and password to Entra ID's /common/oauth2/token endpoint while supplying an arbitrary, unregistered GUID as the client_id parameter instead of a real application ID. Entra ID's differential error responses then leak validity regardless of whether the client_id is legitimate — AADSTS50034 for a non-existent username, AADSTS50126 for a valid username with the wrong password, and, critically, AADSTS700016 ("application not found in directory") when the username and password are both correct, because Entra ID validates the credential before it fails on the unrecognised client. The result is a credential-validity oracle that most defenders misread: AADSTS700016 is ordinarily dismissed as a harmless misconfigured-app error, which is precisely the blind spot both clusters exploited (Help Net Security, 2026-07-13).

The evasion value is in the telemetry: none of these code paths writes a successful sign-in event, and because the client_id is unregistered, the sign-in log entry carries no application name at all — "detections that look for surges against a specific application name may miss this activity entirely, as the field is blank" (Proofpoint, 2026-07-13). Proofpoint attributes two campaigns of opportunistic mass enumeration: UNK_pyreq2323 (January–March 2026, AWS-hosted, 700,000+ distinct spoofed client IDs) and UNK_OutFlareAZ (December 2025–March 2026, Cloudflare-fronted, 3.7M distinct spoofed IDs), whose divergent tooling and client-ID-generation strategies point to parallel invention rather than shared code (The Hacker News, 2026-07-13).

CVE-2026-15409 — SonicWall SMA1000: unauthenticated SSRF (CVSS 10.0) chained to post-auth code injection, actively exploited

SonicWall's PSIRT advisory SNWLID-2026-0008 (first published 2026-07-14) states it has investigated "multiple cases indicating the active exploitation" of two new SMA1000 flaws (SonicWall PSIRT, 2026-07-14); both CVEs carry a same-day CISA KEV listing (recorded in this entry's CVE status, confirmed against the KEV feed). CVE-2026-15409 (CVSS 10.0, CWE-918) is a server-side request forgery in the SMA1000 Work Place interface that lets a remote, unauthenticated attacker force the appliance to issue requests to an attacker-chosen location; the scope-changed CVSS vector (S:C) indicates the SSRF reaches beyond the vulnerable component's own security boundary. CVE-2026-15410 (CVSS 7.2, CWE-94) is a post-authentication code-injection flaw in the SMA1000 Appliance Management Console (AMC) that lets an authenticated administrator-level session run arbitrary OS commands — read together with the pre-auth SSRF, the pair forms a chain from zero access toward root-equivalent appliance control. The affected firmware is SMA1000 6210/7210/8200v on 12.4.3-03245/03387/03434 and 12.5.0-02283/02624/02800; the fix is platform-hotfix 12.4.3-03453 or 12.5.0-02835, and SonicWall explicitly states neither flaw affects SSL-VPN running on SonicWall firewalls or the SMA100 series (SonicWall PSIRT, 2026-07-14).

Microsoft July 2026 Patch Tuesday ships two actively-exploited zero-days — AD FS local EoP (CVE-2026-56155) and unauthenticated SharePoint EoP (CVE-2026-56164)

Microsoft's July 2026 Patch Tuesday is its largest ever by CVE count and carries two zero-days Microsoft confirms were exploited before a fix existed, both added to CISA's Known Exploited Vulnerabilities catalog the same day (BleepingComputer, 2026-07-14). CVE-2026-56155 (CVSS 7.8, CWE-1220) is a local elevation-of-privilege in Active Directory Federation Services: an attacker who already holds a low-privileged authorized session on the AD FS host escalates to administrator (Microsoft MSRC, 2026-07-14). It is a post-foothold escalation rather than an initial-access vector, and Microsoft's advisory credits its own DART incident-response team in the acknowledgements — meaning it surfaced during a live intrusion, so an exposed AD FS host should be treated as a candidate for compromise assessment, not merely patched (Microsoft MSRC, 2026-07-14). CVE-2026-56164 (CVSS 5.3, CWE-306) is the more exposed of the two: a missing-authentication flaw in on-prem SharePoint Server that lets an unauthenticated attacker elevate privileges over the network with no user interaction (Microsoft MSRC, 2026-07-14). Microsoft's mitigation guidance — enable AMSI on the SharePoint/IIS worker processes with Request Body Scan set to Full — indicates the trigger is a crafted HTTP POST body, the same class of exposure this pipeline tracked for the CVE-2026-45659 SharePoint deserialization RCE on 2026-07-02, so operators who already tuned AMSI for that flaw have partial coverage.

FSB Centre 16 (Static Tundra) router-hijacking campaign: 19-agency joint advisory, formal Poland energy-grid attribution and first joint EU/UK cyber sanctions

Background. The FSB Centre 16 network-device cluster is not new — it has a decade-plus public record under the vendor labels Berserk Bear, Energetic Bear, Crouching Yeti, Dragonfly and Ghost Blizzard, and Cisco Talos profiled it in August 2025 as "Static Tundra," documenting long-term compromise of unpatched and end-of-life network gear for configuration theft and persistent collection (Cisco Talos, 2025-08-20). What is new is a same-day trio of actions on 2026-07-13: a much fuller TTP disclosure, a formal government attribution of a destructive attack, and the first coordinated EU/UK cyber-sanctions package.

A joint Cybersecurity Advisory carrying 19 authoring and co-sealing agencies across 13 countries — NSA, CISA, FBI and DC3 (US) alongside the cyber and intelligence authorities of Australia, Canada, New Zealand, the UK, Czech Republic, Denmark, Estonia, Finland, France, Italy, Poland and Sweden — describes FSB Centre 16 opportunistically compromising poorly configured routers across communications, defense industrial base, energy, financial services, government (especially state/local), and healthcare sectors (NSA/CISA/FBI joint advisory, 2026-07-13; NCSC-UK, 2026-07-13). The advisory notes these TTPs overlap with Salt Typhoon activity, so the hardening below counters more than one actor.

The primary access vector is not a novel exploit but weak SNMP hygiene. The actors scan internet IP ranges for SNMP agents that accept common or default community strings, then issue spoofed-source SNMP Set-Requests carrying object identifiers that instruct the device to copy its running configuration to a file (commonly config.bkp or output.txt) and transfer it, usually over TFTP, to a leased VPS or a compromised FTP server (joint advisory, 2026-07-13). The advisory names the exact OIDs abused — 1.3.6.1.4.1.9.9.96.1.1 (Cisco Config Copy) and 1.3.6.1.4.1.9.9.96.1.1.1.1.5 (the destination address for the copied config). A stolen configuration frequently discloses further credentials and additional community strings, feeding lateral movement. Talos's profile records the actor guessing or reusing insecure read-write community strings such as public and anonymous (Cisco Talos, 2025-08-20). Secondarily — "occasionally," per the advisory — the actors exploit known Cisco bugs and the Smart Install (SMI) feature, naming CVE-2018-0171 (the Smart Install pre-auth RCE, in CISA KEV since 2021) and CVE-2008-4128 (end-of-life devices only, no patch). Persistence has historically included the SYNful Knock IOS firmware implant.

The Poland grid attribution. On the same day, the UK together with EU member states formally attributed the destructive 29 December 2025 attack on Poland's energy grid to FSB Centre 16 (NCSC-UK, 2026-07-13). CERT Polska's own incident report describes coordinated destructive activity against 30-plus wind and photovoltaic grid-connection substations — RTU, HMI and protection-relay firmware damaged or system files deleted — and a combined heat-and-power plant serving roughly half a million people, where wiper malware was blocked by the operator's EDR before detonation; CERT Polska tied the activity to the Static Tundra / Berserk Bear / Ghost Blizzard / Dragonfly cluster via VPS, router and anonymizing-infrastructure overlap and called it "the first publicly described destructive activity attributed to this activity cluster" (CERT Polska, 2026-01-30). Note the attribution is contested at the cluster-label level: earlier ESET reporting attributed the same DynoWiper attack to the GRU's Sandworm (BleepingComputer, 2026-01-24), and the EU Council statement names FSB Centre 16 as the parent controlling several groups including Turla — so treat "FSB Centre 16" as an umbrella unit rather than a single team.

The sanctions package is the policy layer: the EU designated 9 individuals and 4 entities and the UK designated 24, covering senior GRU figures, the front company IMPULS accused of recruiting hackers for GRU Unit 29155, Lumma Stealer operators, and the disinformation outlet Rybar LLC (UK Government, 2026-07-13; BleepingComputer, 2026-07-13).

Detection. The telemetry classes to prioritise on network gear: network-flow and firewall logs for inbound SNMP Set-Requests (especially with spoofed or unfamiliar source addresses) and for outbound TFTP sessions initiated from a router/switch management interface to non-management destinations; device syslog and AAA/TACACS+ logs for unexpected "config copy" events, new local-account creation, and unexplained drops in logging volume — Talos documents the actor tampering with TACACS+ configuration to blind logging and standing up GRE tunnels to redirect victim traffic (Cisco Talos, 2025-08-20); and IDS rules keyed to inbound SNMP Set-Requests carrying the config-copy OIDs above, as the advisory recommends (joint advisory, 2026-07-13). Baseline NetFlow for the new GRE tunnel endpoints Talos describes.

Defender takeaway. For a Swiss or European CI operator this is a router-hygiene mandate with a live destructive precedent next door. Disable Smart Install where it is not in active use, confirm CVE-2018-0171 is patched, migrate management SNMP to v3 with authPriv and disable SNMPv1/v2c (or, where legacy SNMP is unavoidable, replace every default/weak community string and enforce read-only), restrict all management protocols to known stations via out-of-band ACLs, use Cisco password hashing type 8 (never 0/4/7), and treat the device configuration held in your management system — not the device itself — as the source of truth so a tampered config is detectable.

Triage: legitimate network-management stations poll SNMP on a predictable cadence from a known IP set, almost always read-only GET/GET-NEXT. The signal is a write (SNMP Set-Request) — particularly one carrying the config-copy OIDs — from a source outside the management range or with an inconsistent/spoofed source address, followed by an outbound TFTP transfer from the device; either alone is weak, the sequence config-write-then-TFTP-egress is the discriminator.

Netherlands NIS2 transposition confirmed: the Senate passed the Cyberbeveiligingswet on 7 July, fixing entry into force at 15 August 2026

UPDATE · originally covered Netherlands NIS2 transposition slips past its 1 July target — Senate vote set for 7 July, entry into force now 15 August 2026 (2026-07-05)

the Dutch NIS2 transposition status this pipeline tracked as "slipped past its 1 July target, Senate vote set for 7 July" has resolved. On 7 July 2026 the Eerste Kamer (First Chamber) passed both the Cyberbeveiligingswet (Cbw, the NIS2 transposition) and the companion Wet weerbaarheid kritieke entiteiten (Wwke, the CER-directive transposition) — the Tweede Kamer had passed them on 15 April — and "de wetten treden op 15 augustus 2026 in werking" ("the laws enter into force on 15 August 2026") (Rijksoverheid.nl, 2026-07-07). The parliamentary vote record confirms broad cross-party support (Eerste Kamer, 2026-07-07). The Cbw covers roughly 8,000 organisations across 18 designated essential/important sectors and imposes a cybersecurity duty of care (including supply-chain risk management), mandatory registration in the NCSC entity register, significant-incident reporting to the relevant CSIRT, and board-level accountability with director training (NCSC-NL).

Why this matters to the constituency: beyond direct applicability to any covered Dutch entity, this is a concrete datapoint for the deployment's standing EU NIS2-transposition watch — a member state moving from indefinite slip to a fixed enforcement date. For Swiss-domiciled organisations with Dutch subsidiaries, NL-incorporated critical suppliers, or cross-border NIS2-equivalent reporting relationships, 15 August 2026 is the operative clock, five weeks out from this brief. The next checkpoint is confirmation the NCSC-NL entity register is live and accepting registrations ahead of the date.

This week's disclosures clustered on third-party, cloud-account and vendor exposure — the breach rarely started inside the victim

Read as a set, the week's confirmed incidents point away from the classic perimeter-RCE story and toward exposure that lives in someone else's account, platform or supply chain.

The third-party / vendor strand: Accenture confirmed a data-theft incident after the handle "888" advertised roughly 35 GB of internal source code (BleepingComputer, 2026-07-08); Deutsche Bank disclosed a third-party-vendor incident after the "Unsafe" ransomware group posted claims (Computing, 2026-07-09); and KDDI named a zero-day in third-party email-platform software as the root cause of a breach affecting about 12 million people (BleepingComputer, 2026-07-09). The cloud-account strand: Nayax, a Bank-of-Lithuania-licensed EEA payment institution, disclosed a cloud-account incident (claimed by "The Syndicate") in its own SEC Form 6-K (Nayax, 2026-07-09); ShinyHunters' Odido (Netherlands telecom) breach drew a Dutch-national-involvement assessment from police voice analysis (Politie, 2026-07-08); and Nextcloud GmbH's own hosting infrastructure exposed roughly 367,000 internal records through a misconfigured public Elasticsearch (Cybernews, 2026-07-10).

Why the pattern matters for the constituency: several victims are directly relevant classes — an EEA-licensed payment institution, an EU telecom, a European cloud vendor — and the shared root cause is exactly the exposure a Swiss/EU public-sector or CI organisation inherits through its suppliers and cloud tenancy. The transferable lesson is that a mature internal patch posture does not cover a vendor's zero-day, a supplier's compromised account, or a misconfigured datastore in your own cloud footprint.

Builds on: 2026-07-08/accenture-confirms-data-theft-888-azure-devops-claim · 2026-07-09/deutsche-bank-unsafe-ransomware-third-party-vendor-incident · 2026-07-09/kddi-isp-email-breach-zero-day-root-cause-update · 2026-07-09/nayax-cloud-account-incident-the-syndicate-claim · 2026-07-10/odido-shinyhunters-vishing-dutch-police-attribution · 2026-07-10/nextcloud-gmbh-elasticsearch-exposure-msb-nrw

GigaWiper: a Golang backdoor that folds a disk wiper, fake-ransomware encryptor and secure-wipe module into one modular implant

Microsoft Threat Intelligence first identified GigaWiper in October 2025 and has now published a code-level analysis of it: a Golang backdoor notable less for any single capability than for its construction — at least three previously separate destructive families folded into one implant as on-demand commands, so an operator can pick the mode of destruction at task time (Microsoft Threat Intelligence, 2026-07-09). The raw-disk wiper command enumerates physical drives over WMI, identifies and spares the Windows installation drive, strips partition metadata from the other drives via DeviceIoControl/IOCTL_DISK_CREATE_DISK, overwrites disk content in 0xA00000-byte chunks (randomising only the first byte of each buffer to dodge naïve all-zero-wipe detections), then forces an immediate reboot. A second command reuses Crucio ransomware code to AES-encrypt files with per-run keys that are never saved and drops no ransom note — destruction wearing an extortion costume — while a third reimplements the C-based FlockWiper in Go for multi-pass secure wiping of the Windows drive. Microsoft ties the families together by code overlap and assesses that the same developer built GigaWiper and Crucio; it withholds actor attribution beyond that lineage. Google's Threat Intelligence Group and Binary Defense track the same activity as BLUERABBIT (Microsoft Threat Intelligence, 2026-07-09; Infosecurity Magazine, 2026-07-10).

Operationally the implant is quieter than its payload. It persists as a scheduled task named OneDrive Update (configured to run roughly every minute and once at startup) and tracks its own execution count in a HKCU\SOFTWARE\OneDrive\Environment registry value, masquerading as Microsoft's sync client. For command-and-control it skips ordinary HTTP: tasking arrives over RabbitMQ/AMQP — a fanout exchange named All for broadcast to every infected client plus a topic exchange for targeted commands — status and output are polled back through a Redis server, and MinIO object storage carries exfiltration, alongside keylogging and screen-capture modules.

CVE-2026-47291 — Windows HTTP.sys pre-auth RCE (CVSS 9.8): ZDI publishes full exploitation mechanics and a detection signature

UPDATE · originally covered CVE-2026-47291 — Microsoft June Patch Tuesday: HTTP.sys pre-auth RCE (CVSS 9.8) headlines the largest-ever release (198 CVEs) (2026-06-10)

CVE-2026-47291 shipped in the June 2026 Patch Tuesday as a headline pre-auth RCE in HTTP.sys, the kernel-mode HTTP driver that terminates HTTP/1.x and TLS for IIS and every service built on the HTTP Server API. The delta is a full exploitation write-up from Zero Day Initiative's TrendAI Research team, published a month after the patch, that documents the precise defect and trigger and materially lowers the weaponisation bar (Zero Day Initiative, 2026-07-10).

The bug is a 16-bit integer overflow in the buffer-reference array that HTTP.sys grows while parsing HTTP/1.x headers: the capacity counter is incremented by five on each growth without a bounds check, and after 13,107 growths it reaches 0xFFFB, so the next increment wraps to 0x0000; the subsequent reference addition then allocates a 40-byte buffer but memmoves roughly 524,256 bytes into it — a ~500 KB kernel-pool heap overflow. Because SChannel delivers each TLS record to the parser as its own buffer, an attacker who places exactly one header line per TLS application-data record establishes a 1:1 record-to-reference correspondence and drives the counter to overflow with a single ~262 KB request. Successful exploitation crashes the box (kernel memory-access exception) and, under favourable pool layout, can execute code in kernel context. Critically, the path is reachable only via HTTP/1.x-over-TLS — HTTP/2 and HTTP/3 use a different parser and are unaffected (Zero Day Initiative, 2026-07-10).

The write-up also corrects the exposure picture the original advisory left fuzzy: the default MaxRequestBytes of 16,384 bytes caps a request at roughly 4,000 header lines — far short of the ~65,536 references needed — so only hosts that raised MaxRequestBytes to at least 262,144 bytes can be driven to the overflow. Microsoft rates the CVE "Exploitation More Likely"; no in-the-wild exploitation is reported as of ZDI's publication (Microsoft MSRC, 2026-06-09).

Zimbra Classic Web Client: crafted-email code execution fixed in ZCS 10.1.19, surfaced by NCSC-CH (no CVE, exploitation unknown)

Zimbra released ZCS 10.1.19 on 2026-07-07 to fix a Classic Web Client issue in which "a specially crafted email could run malicious code when the email is opened," potentially granting access to mailbox information, session data or account settings (Zimbra, 2026-07-07); heise online covered it the same day as a stored cross-site-scripting flaw in the legacy webmail UI (heise online, 2026-07-07). Switzerland's NCSC-CH added the item to its Cyber Security Hub on 2026-07-10, describing it as allowing unauthenticated remote attackers to reach session data, account settings and mailbox contents when a victim opens a malicious email, and explicitly recording the exploitation status as unknown (NCSC-CH / GovCERT.ch, 2026-07-10). Only the Classic Web Client is affected; Zimbra and heise recommend switching users to the Modern Web Client as an interim mitigation.

Forg365: a commercial Microsoft 365 phishing-as-a-service kit bundling device-code + AiTM phishing, in-panel AI lure drafting, and a browser extension for SSO-cookie persistence

ZeroBEC's teardown, corroborated by BleepingComputer and a CSA Labs research note, describes Forg365 as a Telegram-distributed, subscription-priced (5-day trial, $400/month, $3,800/year) Microsoft 365 phishing-as-a-service platform that packages two independent credential-theft paths behind one operator console (ZeroBEC, 2026-07-09; BleepingComputer, 2026-07-09). The device-authorization branch presents a Microsoft-styled verification-code page and drives the legitimate Microsoft Authentication Broker flow; the adversary-in-the-middle branch classifies inbound traffic to decide whether to serve the phishing page or a benign decoy. Both converge on a valid, MFA-satisfied refresh token or session cookie because the victim completes the genuine Microsoft authentication — as CSA Labs puts it, "multifactor authentication does not stop the attack because the victim, not the attacker, is the one completing the MFA challenge" (CSA Labs, 2026-07-10). Two capabilities stand out beyond the already-covered device-code primitive: an AI lure-drafting assistant embedded directly in the panel alongside SMTP rotation, OAuth-app configuration and token vaulting, and ForgCookie — a Chrome/Edge/Brave extension that silently triggers OAuth flows to refresh the stolen SSO cookie so operator access outlives its normal expiry (ZeroBEC, 2026-07-09). ZeroBEC's Entra telemetry tied observed device-code activity to a residential ISP address, with a campaign-linked backend node later performing Microsoft Graph device-registration calls.

Builds on: 2026-07-10/m365-conditional-access-gaps-railway-lshiy-campaigns

'Comment stuffing' — HTML phishing attachments padded to ~2.5 MB to dilute or exhaust AI/NLP email scanners

A SANS Internet Storm Center diary (2026-07-10, Jan Kopriva) dissects a phishing email that presented as a Microsoft Teams/SharePoint document notification and carried a .xls.html double-extension attachment weighing ~2.5 MB — anomalously large for a self-contained HTML page (SANS ISC, 2026-07-10). Decoded from a \uXXXX-escaped document.write() wrapper, the file was ~431 KB, of which only the first ~11 KB was a working SharePoint-themed credential-harvesting page; the rest was a single HTML comment holding roughly 430,000 repeated "X" characters, placed after the functional payload, accounting for ~97% of the file.

The placement rules out the classic goal. Padding after the payload does nothing to conceal the malicious code, and at 2.5 MB the file falls well short of the tens-of-megabytes scan-size limits modern mail security uses, so this is not the MITRE "Binary Padding" scan-size-evasion play. The handler's assessment — explicitly flagged as informed speculation — is that the target is AI/NLP-based content scanning, which a growing number of gateways now run. Citing KnowBe4's earlier "NLP obfuscation" work, the diary notes that "if a message contains enough innocuous material, the weight of the malicious portion can be diluted to the point where the model no longer flags it with sufficient confidence", and that "the same bulk can also make a message large enough so that scanning it using AI-based mechanisms takes too long, leading some solutions to release it rather than delay delivery indefinitely" (SANS ISC, 2026-07-10). The author judges the token-budget-exhaustion goal the more likely of the two here, since a featureless block of one character works as well as crafted filler for that purpose. He is candid that against a well-tuned model the tactic is blunt — "the padding is also about as low-entropy as any data can get, which means it wouldn't help the file blend in with benign content on a statistical level either" — which is precisely why a simple non-AI signature catches it.

ShinyHunters' Odido (NL telecom) breach: Dutch police voice analysis points to Dutch-national involvement; same vishing-into-spoofed-portal playbook, now against an EU telco

Dutch National Police (Team High Tech Crime) announced on 9 July that its investigation into the February 2026 breach of Dutch telecom operator Odido — and its Ben brand — has produced strong indications of Dutch-national involvement, based on forensic voice analysis of a call recorded at the time of the intrusion; police assess the caller as very likely a genuine human speaker (while not fully ruling out synthetic voice) and are publicly appealing for the caller to come forward before the recording is released (Politie, 2026-07-09; NOS, 2026-07-09).

This extends the ShinyHunters vishing-to-spoofed-portal playbook (registry: actor:shinyhunters) already covered in this store to a new victim class — an EU telecommunications operator. The mechanism, confirmed on-record by Odido CEO Tisha van Lammeren, is the same one documented previously: a caller impersonating Odido IT-department staff (T1684.001) used a voice-phishing pretext (T1566.004) to persuade a customer-service employee to log into a spoofed copy of the corporate work environment, harvesting that employee's real credentials (T1078) for the customer-contact system (NOS, 2026-05-12). Odido blocked the account within an hour of noticing the intrusion (NOS, 2026-05-12), but the operators had already bulk-exported 6.2 million customer records (name, address, contact details, customer number, bank account number, date of birth, and passport/driver's-licence numbers) (NOS, 2026-02-12) — the CEO's Dutch quote via NOS: "De hacker wist deze medewerker over te halen om in te loggen op een valse versie van de werkomgeving. Zo heeft hij de inloggegevens van die persoon gestolen" ("the hacker persuaded this employee to log into a fake version of the work environment, and so stole that person's login credentials") (T1213). The Dutch Data Protection Authority has two open investigations — into the adequacy of Odido's customer-system security and into whether it retained former-customer data longer than permitted.

Builds on: 2026-06-26/shinyhunters-used-a-single-vishing-call-into-the-company-s-i

Two 2026 M365 account-takeover campaigns (Railway device-code phishing, LSHIY ROPC spray) beat Conditional Access without breaking MFA

Huntress compared two structurally different but strategically identical 2026 Microsoft 365 account-takeover campaigns, both of which got through tenants whose Conditional Access (CA) policies required MFA — because each used an authentication path CA typically does not inspect (Huntress, 2026-07-09). The "Railway" campaign (March 2026) abused Microsoft's OAuth device-code flow: attackers generate a legitimate device-authorization code, embed it in a lure, and collect the resulting OAuth token (valid up to 90 days) when the victim enters the code at the real Microsoft endpoint — the victim may complete MFA, but the token is already gone, so the flow sidesteps MFA rather than defeating it (T1528). The operation ran from clean Railway.com PaaS IP ranges with trusted reputation (three IPs accounted for ~84% of traffic), used construction-RFP lure themes and in some chains triple-wrapped URLs through Cisco, Trend Micro and Microsoft SafeLinks in sequence, and reached 344 organisations across the US, Canada, Australia, New Zealand and Germany before Huntress published; it was attributed to a commercial phishing-as-a-service operation Huntress tracks as EvilTokens — a subscription platform with a storefront, a support team and AI-assisted lure generation (Huntress, 2026-07-09).

The "LSHIY" campaign (active mid-June 2026) took the opposite approach: no phishing, just 81M+ login attempts from an IPv6 range against Azure CLI using the deprecated Resource Owner Password Credentials (ROPC) OAuth flow, which posts credentials straight to the /token endpoint and never touches the authorization endpoint where most CA policies are enforced (T1110.003, T1078.004, The Hacker News, 2026-07-01). It compromised at least 78 accounts across 64 organisations; the finding that matters for defenders is that 55 of those had active CA policies requiring MFA that failed for predictable scoping reasons (T1556.006): MFA scoped to specific apps such as Admin Portals but not "All Cloud Apps", so Azure CLI slipped through; MFA scoped to specific user groups that omitted the compromised accounts; MFA required only from "untrusted" locations, bypassed by an attacker IP that geolocated inconsistently to the US; and two policies left in report-only mode. Huntress notes one tenant had a CA policy explicitly named "Block Azure CLI" that did not, in fact, block Azure CLI.

CitrixBleed 2 (CVE-2025-5777) weaponised into a repeatable IAB kill chain ending in DragonForce ransomware (STAC3725)

Across the first half of 2026 the Huntress Tactical Response unit worked at least six intrusions at unrelated organisations that reproduced the same seven-step kill chain so faithfully that analysts could predict the next artefact before pulling the log — the basis for their high-confidence assessment that an initial-access broker (IAB) has productised the path from an internet-facing Citrix box to domain-wide encryption, a cluster Sophos independently tracks as STAC3725 (Huntress, 2026-07-09; Sophos X-Ops, 2026-04-16). Initial access is pre-auth exploitation of CitrixBleed 2 (CVE-2025-5777), a memory over-read in NetScaler ADC/Gateway configured as a Gateway or AAA virtual server: a POST to the login endpoint (/p/u/doAuthentication.do and equivalents) with the login form variable present but empty makes the appliance serialise roughly 127 bytes of adjacent process memory into the response, and sprayed at volume this yields live session tokens (T1190, T1550.001). In one reconstructed case a user authenticated normally over LDAP+MFA from a known-good IP at 13:07 UTC; twenty-one minutes later the same session was driven from the attacker's IP with no successful authentication from that IP anywhere in the logs — token replay, with MFA already satisfied and therefore irrelevant (Huntress, 2026-07-09).

The privilege-escalation primitive is what makes the cluster unmistakable, because the hijacked session usually belongs to an unprivileged employee and the operator carries a portable, unsigned LPE tool (dropped to working paths such as C:\temp and renamed per victim — eng.exe, legal.exe, as.exe — often inside a password-protected archive pulled from temp.sh). The tool plants a REG_LINK SymbolicLinkValue under the RdpBus device-class key {28d78fad-5a12-11d1-ae5b-0000f803a8c2} that redirects into the Group Policy state hierarchy (T1112); running gpupdate forces the SYSTEM-context Group Policy engine to write through the planted link into a protected key, and sc start AppMgmt then makes the Service Control Manager relaunch the dropper as NT AUTHORITY\SYSTEM, which creates a backdoor administrator via net user … /add and net localgroup Administrators … /add (T1068, T1136.001, T1098). AppMgmt is chosen because it is always present, normally dormant, and plausibly related to policy processing. Before detonating, the tool snapshots the original key tree and restores it afterwards, leaving the registry indistinguishable from its pre-exploit state to erase the artefacts a responder would key on (T1070). Persistence then rides legitimate remote-management software — ScreenConnect and Zoho Assist, in one case Netbird plus Atera (T1219) — and in the most advanced case the operator used PsExec, Impacket and Mimikatz for lateral movement and credential access (T1003, T1570) before deploying DragonForce ransomware, contained to a single host (T1486). Huntress declines a firm DragonForce-affiliate-versus-IAB attribution given the tactic overlap, and ruled out an alternative NetScaler session-management race-condition flaw because the affected build and the required already-authenticated session to race against did not fit the evidence.

CVE-2026-50656 — Microsoft Defender engine 'RoguePlanet' local privilege escalation now patched; NCSC-CH tracks the ongoing 'Nightmare Eclipse' zero-day series

NCSC-CH's running tracker on the "Nightmare Eclipse" (aka Chaotic Eclipse) researcher's 2026 zero-day PoC series was updated on 2026-07-09 to record that a CVE has been assigned to RoguePlanet: CVE-2026-50656, a local privilege-escalation vulnerability (CWE-59, improper link resolution before file access / "link following") in the Microsoft Malware Protection Engine that underpins Microsoft Defender, System Center Endpoint Protection and Microsoft Security Essentials (NCSC-CH, 2026-07-09). The researcher first disclosed RoguePlanet as an unpatched zero-day on 2026-06-10, describing a race condition in Defender that lets a local attacker "execute arbitrary code or spawn a command shell with SYSTEM-level privileges" (T1068), at which point NCSC-CH logged its status as "Proof of Concept Available, no patch available" (NCSC-CH, 2026-07-09). Microsoft's own record shows the CVE was published 2026-06-16 (CVSS 3.1 7.8, AV:L/AC:L/PR:L/UI:N, rated "Exploitation More Likely", exploitation status "No") and remained without a fix for over three weeks; a revision dated 2026-07-08 confirms Microsoft has now shipped an engine update that closes it — last vulnerable Malware Protection Engine build 1.1.26050.11, first fixed build 1.1.26060.3008 (Microsoft MSRC, 2026-07-08).

Because the Malware Protection Engine (mpengine.dll) auto-updates multiple times a day by default, most estates will already carry the fixed build — Microsoft's guidance is that no manual action is normally required. The operational nuance for this constituency is the exception set: any environment where engine updates are pinned, WSUS-gated, air-gapped, or centrally deferred (System Center Endpoint Protection deployments, offline or OT-adjacent Windows hosts) should explicitly verify the installed engine version rather than assume auto-remediation occurred (Microsoft MSRC, 2026-07-08). Microsoft also notes that hosts with Defender disabled are not in an exploitable state even though vulnerability scanners flag the on-disk binaries. Triage: the exploit abuses a symlink/junction race against files Defender is actively scanning, so the telemetry class is symlink/junction creation targeting Defender scan paths and, on success, MsMpEng.exe (the engine's scan host) spawning an unexpected child process with a SYSTEM token outside the normal signature/engine-update cadence — the update-cadence anchoring is the discriminator from routine engine activity. This closes RoguePlanet specifically; NCSC-CH continues to track the wider Nightmare Eclipse PoC series as a home-region authority, which is the reason a single-host LPE closure like this one is worth surfacing to this constituency at all.

KDDI names the root cause of its ISP email-platform breach: a zero-day in third-party software the vendor had not recognized

UPDATE · originally covered KDDI third-party email platform breach exposes up to 14.22 million credentials across six Japanese ISPs (2026-06-29)

KDDI's 6 July update — reported by BleepingComputer on 8 July — discloses the confirmed root cause and exact scale of the breach of the shared email platform serving STNet, JCOM, Chubu Telecommunications, NIFTY and BIGLOBE. The platform was compromised on 16 May 2026 via a zero-day vulnerability in an (still unnamed) third-party software component — a flaw that, per KDDI, "was not recognized by the software vendor" as of KDDI's 17 June confirmation date and which the vendor is now reporting to public authorities (BleepingComputer, 2026-07-08). KDDI confirmed final counts of 12,233,087 exposed email addresses and 7,616,173 exposed passwords — down from the earlier "up to 14.22 million" estimate (BleepingComputer, 2026-06-28) — deployed EDR post-incident, completed a forensic audit on 23 June confirming the flaw was patched with no other issues remaining, and notified Japan's Personal Information Protection Commission and the Ministry of Internal Affairs and Communications.

Neither report names the exploited third-party product; KDDI has stated only "third-party software", and that ambiguity is in the source, not omitted here.

Nozomi documents two new Golang IoT/Linux DDoS botnets (Apex2, c2c/meow) built for speed and reuse over sophistication

Nozomi Networks Labs' AI-assisted honeypot triage flagged two Golang-based DDoS botnet samples this spring that stand out from the routine volume of Mirai-derived variants: Apex2 and c2c (distributed under the filename "meow") (Nozomi Networks Labs, 2026-07-06). Apex2 is a direct structural evolution of the earlier Apex botnet: infection begins with Telnet connections and credential brute-forcing, followed by download-and-execute of the Golang payload, which registers with its C2 over a plaintext protocol (host OS/architecture) and ships builds for Linux (arm, arm64, mipsle, ppc64) and Windows (386, amd64). Its named flood commands include cf (an HTTP(S) flood specifically tuned to bypass Cloudflare via randomized User-Agent lists and long keep-alive timeouts), udp/pps, discord/game UDP floods, and three TLS-flood variants (tls, tlsplus, tlsplusbypass). c2c/meow is architecturally simpler — a Golang flooder with no built-in propagation (a separate SSH scanner handles brute-forcing and delivery) that authenticates to a hardcoded C2 over plaintext JSON-over-TCP, checks for passwordless sudo (sudo -n true) to self-escalate, then persists by copying itself to /usr/local/bin/cpufreqd and registering a fake systemd unit masquerading as a "CPU Frequency Daemon" — supporting ten flood-module types (icmp, dnsudp, udp, http, directhttp, fasthttp, betterhttp, tcp, tcphandshake, dnstcp).

Nozomi's stated point for defenders is that neither family is sophisticated — both lean on commodity Golang tooling, weak/default credentials and exposed Telnet/SSH interfaces rather than novel exploitation — and that the lack of sophistication does not reduce the risk at scale, because the build-and-deploy cycle for such botnets is getting faster. ATT&CK mapping: T1110 Brute Force (Telnet/SSH), T1105 Ingress Tool Transfer, T1548.003 Abuse Elevation Control Mechanism: Sudo (c2c's passwordless-sudo self-escalation), T1543.002 Create or Modify System Process: Systemd Service with T1036.005 Masquerading (the fake cpufreqd unit), and T1498 Network Denial of Service for the flood modules.

Sygnia: an AI-orchestrated AWS intrusion reached broad compromise in ~72 hours — four keys from four accounts used from one source in the same second

Sygnia's incident-response investigation of a financially-motivated AWS cloud intrusion found no novel malware or zero-day — every individual technique maps to a long-tracked MITRE ATT&CK ID — but the operationalisation was materially faster than typical manual intrusions, which Sygnia attributes to AI-assisted or agentic tooling (Sygnia, 2026-07-08). After obtaining an initial access key via a weakness in an internet-facing application, the actor ran four workstreams in parallel — secrets theft (ECS/EC2 environment variables, GitHub/Bitbucket CI/CD runner env vars, S3 plaintext secrets, Secrets Manager, SSM Parameter Store); persistence (new IAM users, EC2/ECS reverse shells, modified deployment files); RDS exfiltration via several hundred distinct SQL queries across dozens of databases; and reversible impact (S3 access denial, ECS scaled to zero, SQS purges) used purely as extortion leverage — and repeated the full playbook on every newly obtained credential rather than progressing linearly. The most striking artefact: four different AWS access keys from four separate accounts were used from the same source IP and user-agent within a single observed second, which Sygnia assesses is very hard to explain as manual operation (Sygnia, 2026-07-08).

Scripts, structured reporting output, and commit messages/branch names framing the activity as an authorized "pentest"/"red team" with a fabricated CEO sign-off are consistent with LLM-generated tooling — possibly including prompt-framing meant to reduce refusal from AI assistants being abused by the operator. Sygnia maps the case onto the same tactic distribution (Execution, Discovery, Credential Access, Collection, Defense Evasion) that Anthropic's June 2026 LLM ATT&CK research found concentrated in banned AI-abuse accounts. Relevant IDs per Sygnia include T1651 Cloud Administration Command, T1552/T1528 (credential/token harvesting), T1087/T1580/T1619 (account/cloud-infra/storage discovery re-run per key), T1578 (modify cloud compute infra) and T1078 Valid Accounts.

Mandiant "Ghost in the Database": recovering an active ADFS token-signing key from Machine DPAPI when the WID/DKM Golden SAML path fails

Background. Golden SAML — forging SAML assertions by stealing an identity provider's token-signing key — has been public tradecraft since CyberArk's 2017 disclosure (CyberArk, 2017), and Mandiant previously documented network-based extraction of ADFS secrets during the UNC2452/SolarWinds intrusions (Mandiant). The standard extraction path pulls the encrypted signing key from the ADFS Windows Internal Database (WID) and decrypts it with Distributed Key Manager (DKM) material stored in Active Directory. This new Mandiant write-up documents a variant that defeats that assumption when ADFS configuration has drifted.

During a red-team engagement, Mandiant found that ADFS deployments with AutoCertificateRollover disabled (Get-AdfsProperties → AutoCertificateRollover: False) and certificates rotated manually can leave the WID configuration database holding only a stale "ghost" certificate record, while the ADFS service actually signs tokens with a newer certificate whose private key lives in the machine CAPI store (Mandiant, 2026-07-07). In that state the classic path still "works" mechanically — the WID blob decrypts via DKM — but Entra ID rejects the resulting token with AADSTS500172 because the key is no longer the one in use.

The key's real location and protection. The active private key sits under C:\ProgramData\Microsoft\Crypto\RSA\MachineKeys\, with the certificate enrolled in the LocalMachine\My store. It is protected by Machine DPAPI (not user-bound DPAPI): the DPAPI_SYSTEM LSA secret plus machine masterkeys under the S-1-5-18 (SYSTEM) context at C:\Windows\System32\Microsoft\Protect\S-1-5-18\. Machine-scoping is deliberate — it keeps the key usable across service-account password changes, gMSA rotations and reboots — but it also means a SYSTEM-level actor can recover the key entirely from the host. Mandiant confirmed recovery with SharpDPAPI /machine, which enumerated the active key material under that path (the CNG Crypto\Keys store was not in use in the assessed environment) — no interaction with the live ADFS process or LSASS is required, reducing visibility for defenses that watch only credential-dumping/process-memory access (Mandiant, 2026-07-07).

Kill chain (ATT&CK). SYSTEM-level foothold on the ADFS host → recover Machine-DPAPI-protected masterkeys and the CAPI signing key (T1552 Unsecured Credentials, via SharpDPAPI /machine) → forge a SAML assertion impersonating a Global Administrator (T1606.002 Forge Web Credentials: SAML Tokens) → Entra ID accepts it as a valid federated authentication assertion, yielding Global Administrator access to the Microsoft 365 tenant with MFA and conditional access fully bypassed (T1078.004 Valid Accounts: Cloud Accounts). Because the forged assertion is honoured for all SAML relying-party trusts, the blast radius extends to every SaaS platform federated through the same ADFS, not just Microsoft services.

Hunt and detection. The drift condition itself is observable: ADFS Event ID 385 fires when the WID record and the actively-used signing certificate diverge, and self-resolves only once AutoCertificateRollover is re-enabled and a rollover runs. For key-theft detection, Mandiant recommends SACL-based object-access auditing (Security Event ID 4663) on C:\ProgramData\Microsoft\Crypto\RSA\MachineKeys\ and C:\Windows\System32\Microsoft\Protect\S-1-5-18\, treated as correlation evidence rather than a standalone signal. The strongest analytic is cross-source: correlate ADFS token-issuance/claims events (Event IDs 299 and the 1200-series, version-dependent) against Entra ID sign-in logs to surface federated sign-ins with no matching upstream authentication context, baselining claim sets, IP ranges and user-agents per relying-party trust for privileged accounts — neither log source alone is sufficient (Mandiant, 2026-07-07).

Hardening. Migrate token-signing certificates to an HSM to eliminate the software-accessible key and thus the Machine DPAPI extraction path entirely; run ADFS under gMSA to reduce manual-rotation drift; govern ADFS servers as Tier 0 (restricted admin paths, dedicated PAWs, separation from general server administration). When AutoCertificateRollover is disabled, a manual rotation must include Set-AdfsCertificate — installing the certificate alone is insufficient — and be validated with Get-AdfsCertificate; a subsequent Event ID 385 signals lingering inconsistency. Organisations migrating to native OIDC federation remove this attack path altogether (Mandiant, 2026-07-07). ADFS remains widely deployed for on-prem/hybrid identity across Swiss and EU public-sector estates mid-migration to Entra ID, making this a direct Tier 0 hardening item for the constituency.

Git commit-signature malleability mints a second "Verified" GitHub commit with a different hash — defeating hash-based blocklists

Jacob Ginesin (Carnegie Mellon PhD student, Cure53 auditor) published research on 2 July, amplified by The Hacker News on 8 July, showing that Git/GitHub's "Verified" commit badge is not a unique identifier: given any signed commit, an attacker without the signing key can mint a second, distinct commit with an identical tree, identical author/date metadata, and a valid signature that still shows "Verified" — differing only in its resulting hash (The Hacker News, 2026-07-08; Ginesin, arXiv, 2026-07-02). The root cause is signature malleability, not a hash collision. A commit's SHA is computed over everything inside it, including the raw signature bytes in its header, and many signatures can be rewritten into a different-but-valid form.

Three malleation routes are demonstrated: (1) for ECDSA, the classical algebraic symmetry that turns a valid pair (r,s) into (r, n−s) using only public curve parameters, producing a second equally-valid signature over the same payload with different bytes and therefore a different commit hash; (2) for RSA and EdDSA under OpenPGP, appending an ignorable experimental subpacket in the unhashed subpacket region defined in RFC 4880 §5.2.3; (3) an analogous X.509/S-MIME path. GitHub does not normalize or canonicalize a signature before verifying it — no strict encoding enforcement on S/MIME, no stripping of the manipulable OpenPGP fields, and non-canonical ECDSA values accepted as-is (The Hacker News, 2026-07-08). A public exploitation tool implementing all three attacks, plus demo repos where the malleated commits still show "Verified", is released (Ginesin, git-chain-malleator). Ginesin reported to GNU/Git in January and GitHub in March 2026; neither had shipped a fix at publication, and no CVE is assigned. Maps to T1195.002 Compromise Software Supply Chain as a control-bypass primitive.

GhostApproval (CVE-2026-12958, CVE-2026-50549) — symlink + confirmation-UI misrepresentation lets a malicious repo write outside six AI coding assistants' workspace sandbox

Wiz Research published GhostApproval on 8 July, a systematic vulnerability pattern combining CWE-61 (symbolic-link following) with CWE-451 (UI misrepresentation of critical information) found, in varying severity, across six AI coding assistants: Amazon Q Developer, Cursor, Google Antigravity, Augment, Cognition Labs' Windsurf and Anthropic's Claude Code (Wiz Research, 2026-07-08). A malicious repository plants a symlink inside the workspace that resolves to a sensitive path outside it — e.g. a file named project_settings.json that is actually a link to ~/.ssh/authorized_keys — then a README or prompt instructs the agent to "update" the file. In several tools the agent's own reasoning identifies the true target, yet the confirmation dialog still shows the harmless in-workspace name, so the user rubber-stamps a write to the real target, enabling persistent passwordless SSH access or other host compromise. Windsurf exhibited a pre-authorization write — the file was modified on disk before the Accept/Reject buttons even rendered, making the prompt an "undo" button rather than a gate.

AWS assigned CVE-2026-12958 (missing symlink validation in Language Servers for AWS, CVSS 8.5, fixed in language-servers 1.69.0 / @aws/lsp-codewhisperer 0.0.117) and Cursor assigned CVE-2026-50549 (sandbox escape via symlink + failed path canonicalization, fixed in Cursor 3.0), both confirming arbitrary out-of-workspace file write as the impact (AWS GHSA-6v3r-4p5c-mrp5, 2026-06-23; Cursor GHSA-3v8f-48vw-3mjx, 2026-06-05). Google fixed Antigravity (CVE pending at publication). Augment and Windsurf acknowledged the report but were still testable-vulnerable at disclosure (Wiz Research, 2026-07-08). Anthropic assessed the report as outside its threat model for Claude Code — its stated rationale is that a user who starts a session in a directory has already extended trust to it — while noting it had shipped a symlink warning in the Edit/Write permission dialog in v2.1.32 (5 Feb 2026) as unrelated proactive hardening (Wiz Research, 2026-07-08). Mapped to T1195.002 Compromise Software Supply Chain, T1222 File and Directory Permissions Modification (via symlink) and T1552.004 Unsecured Credentials (authorized_keys write).

ESET Threat Report H1 2026: first Android malware using generative AI at runtime, ClickFix detections more than double, record QR-phishing, 100+ EDR-killers

ESET's semi-annual threat-landscape report (telemetry December 2025–May 2026) flags four developments a Tier 2/3 team should track (ESET / WeLiveSecurity, 2026-07-08; ESET press release, 2026-07-08).

First, ESET analysed roughly 900,000 "AI skills" — small functional components used by AI agents — and found tens of thousands suspicious and thousands outright malicious, an expanding attack surface in the emerging agentic-AI ecosystem. Second, it identified PromptSpy, described as the first known Android malware to use generative AI (specifically Google's Gemini) inside its own execution flow to interpret UI elements and adapt behaviour across devices at runtime rather than relying on hardcoded logic — following the first AI-powered ransomware disclosed in 2025 (ESET, 2026-07-08). Third, ClickFix (the fake-error social-engineering technique) has expanded beyond fake CAPTCHA prompts into AI-themed help pages, browser extensions and cloud-authentication scenarios, with ESET detections more than doubling between H2 2025 and H1 2026. Fourth, QR-code phishing ("quishing") reached record levels, with roughly 11% of all ESET-detected phishing emails in H1 2026 using QR codes to move victim interaction onto mobile devices and evade cursory inspection. Ransomware activity continued unabated with over 100 distinct EDR-killer tools now catalogued by ESET, though a declining share of victims are reportedly paying.

CVE-2026-53359 — Linux KVM/x86 "Januscape": shadow-MMU use-after-free enables guest-to-host VM escape on Intel and AMD

Researcher Hyunwoo Kim (V4bel) disclosed Januscape (CVE-2026-53359), a use-after-free in the shadow-MMU emulation of KVM/x86 (arch/x86/kvm/mmu/mmu.c) whose root cause traces to a 2010 commit — roughly 16 years dormant before the fix landed upstream on 16 June 2026 (BleepingComputer, 2026-07-07). The bug fires when kvm_mmu_get_child_sp() reuses a shadow page without comparing its role, producing a mismatched direct/indirect flag and an incorrect GFN computation; orphaned rmap entries survive memslot deletion and are later dereferenced after the backing memory is freed (V4bel, 2026-07-07). The upstream fix is commit 81ccda30b4e8 (16 June 2026); the researcher gives the vulnerable range as commit 2032a93d66fa (2010-08-01) through that fix (kernel.org, 2026-06-16; V4bel, 2026-07-07).

This is a genuine guest-to-host escape: a root user inside a KVM guest can trigger the UAF from purely guest-side actions, on both Intel and AMD hosts — the researcher calls it "the first guest-to-host exploit research triggerable on both" vendors (V4bel, 2026-07-07). Januscape was submitted as a live 0-day against Google's kvmCTF program. A public PoC that panics the host kernel — a denial of service against every co-tenant on the same physical host — is released; a full working host-compromise/RCE exploit exists but the researcher is deliberately withholding it (BleepingComputer, 2026-07-07). Mapped to T1611 Escape to Host.

CVE-2026-48614 — Plesk XML API code injection: authenticated low-privilege user to root (CVSS 9.9)

The Centre for Cybersecurity Belgium (CCB) published a standalone "patch immediately" advisory on 8 July for CVE-2026-48614, a CWE-94 (improper control of code generation / code injection) flaw in Plesk's XML API (CCB, 2026-07-08). An authenticated, low-privilege panel user can send a crafted XML-API request that bypasses the intended authorization boundary and injects arbitrary configuration directives into upstream config generation; because input neutralisation is broken, this yields an arbitrary file write performed as root, i.e. local privilege escalation from any authenticated panel account to full root on the hosting server. CCB scores it CVSS 3.1 9.9 (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H) (CCB, 2026-07-08). Plesk's own advisory confirms the CVE and the LPE impact, thanks independent researcher Georgii Shutiaev for the disclosure, and lists affected versions below 18.0.30, patched in 18.0.30 through 18.0.78.4, with 18.0.79 and later unaffected (Plesk, 2026-07-03). Neither CCB nor Plesk reports in-the-wild exploitation at publication. Mapped to T1068 Exploitation for Privilege Escalation.

The prerequisite is only a valid low-privilege authenticated session — and that is the point for defenders: on multi-tenant shared-hosting Plesk installs, every hosting customer already holds such an account, so the flaw collapses tenant isolation and turns any customer into a path to root on the shared server and thus to every co-tenant's sites and data. Plesk is broadly deployed across Swiss and EU web-hosting providers and public-sector/SME web infrastructure, which is why CCB — a national authority (Admiralty A) — escalated it rather than leaving it to routine patching.

Check Point: Iran MOIS-linked "Cavern Manticore" ships a modular .NET C2 that uses three compilation formats as an anti-analysis layer, delivered via SysAid RMM abuse

Check Point Research documented Cavern Manticore, an Iran MOIS-linked APT it assesses shares technical and infrastructure overlap with MuddyWater and OilRig's Lyceum subgroup, targeting Israeli government and IT-sector organisations (Check Point Research, 2026-07-06). Its namesake framework, Cavern, is a modular post-exploitation .NET C2 whose components are deliberately compiled into three different binary formats: pure IL-only .NET (the mhm.dll file-ops/DPAPI-decrypt module, db.dll SQL browser, ode.dll LDAP/AD-recon module), Mixed-Mode C++/CLI IL+native (the uxtheme.dll Cavern Agent core), and .NET 8 NativeAOT native-only (n-HTCommp.dll HTTPS/WebSocket transport, n-ten.dll network recon/SMB brute-force, n-sws.dll SOCKS5/WSS tunnel). The compilation-format diversity is itself the anti-analysis layer: each format demands a different reverse-engineering toolchain, and NativeAOT strips framework symbols and resolves security-sensitive P/Invoke calls (WNetAddConnection2, NetShareEnum, NetLocalGroupGetMembers) through runtime descriptor tables rather than the PE import table, hiding capability from import-based triage (Check Point Research, 2026-07-06).

Delivery is the transferable part: the actor abused SysAid's legitimate software-update/deployment feature — not a SysAid vulnerability — to push a WinDirStat DLL-sideloading package that loads the trojanized uxtheme.dll as the Cavern Agent, which exports 83 functions mimicking the real Windows theming library (82 empty stubs; the one live export, EnableThemeDialogTexture, is the C2 entry point) — a sandbox trap for automated analysis that only invokes default exports. Each loaded module is isolated in its own .NET AppDomain via a MarshalByRefObject proxy so modules can be unloaded cleanly after use, leaving minimal forensic residue; most samples score zero or near-zero on VirusTotal. ATT&CK: T1574.002 DLL Side-Loading, T1027 Obfuscated Files or Information (via compilation-format diversity), T1620 Reflective Code Loading (AppDomain-isolated modules), T1219 Remote Access Software (SysAid deployment abuse).

Ubiquiti UniFi SAB-066 — 25 vulnerabilities incl. unauthenticated CVSS 10.0 command injection in UniFi Connect (CVE-2026-50746)

NCSC-NL published advisory NCSC-2026-0221 on 7 July 2026 covering Ubiquiti's Security Advisory Bulletin 066 (vendor-published 2026-07-02): 25 vulnerabilities spanning the UniFi Connect, Talk, Access, Network and Protect applications plus the UniFi OS platform itself across the Dream Machine / Cloud Gateway / Cloud Key / Network-Video-Recorder / Enterprise-Fortress-Gateway hardware families (NCSC-NL, 2026-07-07). This is a distinct, larger disclosure from the CVE-2026-34908/-34909/-34910 UniFi OS chain covered on 2026-06-24 — different CVEs, broader scope. The most severe, CVE-2026-50746 (CVSS 10.0), is an improper-access-control flaw in UniFi Connect (< 3.4.20) letting a network-adjacent unauthenticated attacker execute OS command injection on the host device; CVE-2026-50747 (CVSS 9.9, authenticated SQLi in Talk), CVE-2026-50748 (CVSS 9.9, command injection in Access), CVE-2026-54402 (CVSS 9.9, command injection in UniFi OS) and CVE-2026-55115 (CVSS 9.9, SSRF in Protect) round out the critical set, and CVE-2026-54403 (CVSS 8.6, path traversal in UniFi OS) bypasses authentication outright and is explicitly flagged by Ubiquiti as chainable to drop the low-privilege prerequisite of the others. SOCRadar confirms no functional public PoC and no confirmed in-the-wild exploitation as of 2026-07-08 (SOCRadar, 2026-07-08).

Cisco Talos: China-nexus UAT-7810 expands its ORB network with LONGLEASH/DOGLEASH/JARLEASH via unpatched Ruckus and ASUS routers

Cisco Talos profiled UAT-7810, a China-nexus actor Talos assesses with high confidence is tasked with building and maintaining Operational Relay Box (ORB) networks — relay/proxy infrastructure built from compromised networking gear that secondary China-nexus APTs use to launder the origin of operations against high-value targets (Cisco Talos, 2026-07-07). Talos names UAT-5918 — previously documented targeting Taiwanese critical infrastructure — as one such downstream consumer. Initial access is exploitation of known, unpatched vulnerabilities in Ruckus wireless routers (CVE-2020-22653, CVE-2020-22658, CVE-2023-25717) and ASUS AiCloud routers (CVE-2025-2492), a tactic UAT-7810 has used since 2025 rather than a fresh zero-day (T1190). The malware suite, internally "ff-agent", now includes LONGLEASH — an enhanced successor to the SHORTLEASH backdoor adding reverse-shell and HTTP/DNS/SOCKS/TCP/ICMP/UDP multi-protocol proxying (T1090.003) — plus DOGLEASH, a passive C-based Linux backdoor, and JARLEASH, a Java-based admin tool for file management and FTP/SFTP access; it is built with Boost.Asio, custom protobuf encoding and MbedTLS TLS proxying, compiled for MIPS/ARM/x64, and self-deletes if tampering or a suspicious connection is detected (T1070).

Netherlands NIS2 transposition slips past its 1 July target — Senate vote set for 7 July, entry into force now 15 August 2026

UPDATE · originally covered Netherlands NIS2 (Cyberbeveiligingswet) clears the lower house — entry into force targeted for 1 July 2026 (2026-06-29)

the Dutch NIS2 transposition — the Cyberbeveiligingswet (Cbw) plus the companion Wet weerbaarheid kritieke entiteiten — has missed the 1 July 2026 entry-into-force target the prior weekly reported as the government's goal.

The Eerste Kamer (Senate) tabled its government response to the second committee report ("nota naar aanleiding van het tweede verslag") on 29 June 2026 — the last written-preparation step before plenary debate — and the Senate's own bill-tracking page now states the floor vote will take place on 7 July 2026, noting the bill was adopted by the Tweede Kamer on 15 April 2026 (Eerste Kamer, bill 36764). iBestuur reports the government's revised entry-into-force target is now 15 August 2026, roughly six weeks later than previously communicated (iBestuur, 2026-07-01).

The substantive scope is unchanged from prior coverage: NCSC-NL as designated supervisor, a three-step 24h/72h/one-month incident-notification protocol, essential-entity fines up to EUR 10M or 2% of global turnover, personal board liability for security-measure oversight, and an expansion of in-scope Dutch entities from roughly 1,000 to roughly 8,000. This is the fourth documented slip in the Dutch NIS2 timetable (originally targeted Q3 2025).

Kemp LoadMaster CVE-2026-8037 — exploitation attempts confirmed the day the PoC dropped

UPDATE · originally covered CVE-2026-8037 — Progress Kemp LoadMaster: pre-auth RCE via uninitialized heap in the /accessv2 API (2026-06-30)

eSentire's Threat Response Unit reports that in-the-wild exploitation attempts against CVE-2026-8037 — the Progress Kemp LoadMaster pre-auth OS command-injection flaw reachable through the /accessv2 API endpoint (CVSS 9.6–9.8) — began 2026-06-29, the same day a public proof-of-concept was released, confirming the compressed PoC-to-exploitation timeline (eSentire TRU, 2026-06-30).

The observed attempts were unsuccessful, with no post-compromise activity, but eSentire assesses that public PoC availability plus detailed technical write-ups will drive continued and likely more successful attacks near-term (The Hacker News, 2026-07-01). Affected versions remain LoadMaster 7.2.63.1 and earlier (GA) and 7.2.54.17 and earlier (LTSF); Progress shipped patched firmware in early June 2026. Patch remains the primary mitigation; disabling the LoadMaster API where not required removes the /accessv2 attack surface entirely. Hunt /accessv2 traffic for malformed/oversized parameters and repeated probing from related sources in a short window (T1190 → T1059).

CVE-2026-8037 — Progress Kemp LoadMaster: pre-auth RCE via uninitialized heap in the /accessv2 API

CVE-2026-8037 (CVSS 9.8) is a pre-authentication RCE in Progress Kemp LoadMaster, an edge load balancer (watchTowr Labs, 2026-06-29 · Trend Micro ZDI, 2026-06-09). The escape_quotes() function in the access executable allocates buffers via uninitialized malloc() without null-terminating escaped strings; a sprayed JSON payload to /accessv2 (four single-quotes expanding to 16 bytes) overwrites heap metadata in adjacent freed chunks, and the subsequent __sprintf_chk() reads out-of-bounds into attacker-controlled data, reaching code execution as root with no authentication. watchTowr published the full mechanics. Affected: GA ≤ 7.2.63.1 and LTSF ≤ 7.2.54.17; fixed in v7.2.63.2 (which switches to calloc() with proper null termination). A second bulletin CVE, CVE-2026-33691, bypasses file-upload extension checks via OWASP CRS whitespace padding. Progress reports no known active exploitation. Hardening: patch to v7.2.63.2 and restrict the management interface to a dedicated admin VLAN; perimeter anomaly detection for unusual character sequences in JSON POSTs to /accessv2.

KDDI third-party email platform breach exposes up to 14.22 million credentials across six Japanese ISPs

Japanese carrier KDDI disclosed that a threat actor exploited a vulnerability in third-party software integrated into its centralised ISP email-management platform, with unauthorised access detected on approximately 2026-06-17 (BleepingComputer, 2026-06-28). The breach potentially exposed email addresses and passwords for up to 14.22 million subscriber accounts across six ISPs running on the shared platform — STNet, JCOM, Chubu Telecommunications, Nifty, Biglobe and a further KDDI ISP; KDDI states some passwords were stored hashed or encrypted and that 14.22 million is a worst-case figure pending forensic completion (SecurityAffairs, 2026-06-28; Infosecurity Magazine, 2026-06-24). No CVE for the third-party software flaw and no threat actor have been named; KDDI notified Japan's Personal Information Protection Commission and advised affected users to change passwords and enable MFA.

Why it matters to us: The structural lesson, not the jurisdiction, is the signal — a single vulnerable dependency in a shared multi-tenant email-management plane produced a six-ISP blast radius, the same exposure model any European telco or managed-ISP operator carries when subscriber-mail administration is consolidated onto one vendor platform. The immediate downstream risk for Swiss/EU defenders is credential-stuffing: 14.22 million leaked email/password pairs will surface in combolists and feed phishing-as-initial-access. Hunt for anomalous authentication against external-facing services from Japanese-ISP email address spaces, and treat any reused-password exposure on those domains as a stuffing precursor. Inventory third-party vendor access to your own subscriber/identity-management platforms and enforce MFA on the administration plane itself.

FortiBleed

The W25 top story continued without a scale revision — the device count holds at the 86,644 figure the dailies reported — but the in-window development is the clearest state-interest signal yet: CISA updated its hardening alert on 06-22 to link Fortinet's revised guidance, and reporting now confirms that on in mid-June the Russian-speaking operator completed offline Kerberos-hash cracking from captured FortiGate configs and immediately exfiltrated DFS backup data from a NATO-aligned defence contractor — a full AD domain takeover (Security Affairs). Outstanding for defenders: treat any FortiGate admin/VPN credential active May–June 2026 as compromised, rotate, then hunt AD for pass-the-hash, DCSync and DFS-backup exfiltration (Kerberos ticket anomalies, LSASS access, ntdsutil/impacket artefacts). Patch level is irrelevant — this is credential reuse, not a new CVE.

CVE-2026-20245 — Cisco Catalyst SD-WAN Manager: Mandiant reconstructs the full zero-day chain

Mandiant (GTIG) published the first complete TTP chain on 06-24 for the Catalyst SD-WAN Manager zero-day activity, observed at a service provider: a peering/authentication bypass (CVE-2026-20127, CVE-2026-20182) leading to credential manipulation, then local privilege escalation to root via a malicious CSV upload (CVE-2026-20245) to plant a root backdoor. NCSC-CH posted on it, giving it direct Swiss relevance. Telco and public-sector SD-WAN operators should hunt for unexpected file writes under the web-UI service account and root-owned artefacts post-dating the patch.

Mandiant documents the full Cisco Catalyst SD-WAN exploitation chain — CSV-injection to a root backdoor

UPDATE · originally covered Mandiant publishes the forensic reconstruction behind Cisco SD-WAN Manager CVE-2026-20245 (2026-06-26)

Google Mandiant (GTIG) published (2026-06-24) the first complete TTP chain for the Cisco Catalyst SD-WAN Manager zero-day activity, observed at a service-provider victim from late 2025 into 2026 (Google Mandiant, 2026-06-24). NCSC-CH amended its Security Hub post to add the report on 2026-06-25 (NCSC-CH Security Hub post 12579).

The chain: authentication bypass via CVE-2026-20182/CVE-2026-20127 (rogue peering connection), then privilege escalation via CVE-2026-20245 — a malicious evil_tenant.csv uploaded through the request tenant-upload CLI carries unsanitised shell commands that append a troot root user to /etc/passwd and /etc/shadow, after which the actor reverts configuration changes and deletes the file for anti-forensics. This gives defenders concrete hunts the earlier advisory could not: search SD-WAN Manager instances for unexpected /etc/passwd additions, evil_tenant.csv artefacts, and request tenant-upload execution in CLI logs.

Cisco Catalyst SD-WAN Manager CVE-2026-20245

Mandiant's Google Threat Intelligence Group published a forensic reconstruction of an intrusion in which Cisco Catalyst SD-WAN Manager (formerly vManage) was compromised through CVE-2026-20245 as a zero-day — exploited at a communications service provider from late 2025 through March 2026, months before Cisco's advisory (Mandiant/GTIG, 2026-06-24). Mandiant attributes the activity to no named actor. The reason this matters beyond one victim: SD-WAN Manager is the control plane for an entire WAN fabric — root on the controller is push-access to every managed edge device — so it warrants the same monitoring tier as a VPN concentrator or firewall, and it is now one of several Cisco SD-WAN flaws confirmed exploited during 2026.

The vulnerability. CVE-2026-20245 (CVSS 7.8, no workaround) is a command-injection weakness in the SD-WAN Manager CLI tenant-upload handler: the feature that ingests a tenant-list CSV fails to sanitise file content before it reaches a shell context, so an authenticated operator can embed OS commands inside a crafted CSV and have them execute as root on the underlying Linux host (Cisco PSIRT, cisco-sa-sdwan-privesc-4uxFrdzx). The injected commands appended a new UID-0 account (troot) to the host's local account databases, giving the actor a persistent root login independent of the vManage application's own user model.

Kill chain (as Mandiant documents it):

• Initial access — the actor reached an authenticated position by abusing peering-authentication-bypass flaws CVE-2026-20127 / CVE-2026-20182 to enrol unauthorised peering and obtain SSH as the vmanage-admin account, or alternatively by using certificate material stolen in a previous compromise (T1190, T1078.004).
• Privilege escalation — exploitation of CVE-2026-20245 via the crafted tenant CSV, executing as root (T1068).
• Persistence — creation of the troot UID-0 account in the host account databases, reachable via su (T1136.001).
• Defense evasion / anti-forensics — the actor changed the legitimate admin password and then reverted it to its original value to reduce detection probability, and deleted command history, syslog entries, and the uploaded files after use (T1070.003).

Hunt and detection concepts. The decisive gap is that vManage's own health dashboards do not surface OS-level account creation — detection has to happen on the underlying host. Baseline and monitor /etc/passwd and /etc/shadow for accounts added since a known-good snapshot (a UID-0 account other than root is the high-fidelity signal here). Review SD-WAN Manager audit logs for tenant-upload CLI/API invocations and correlate them with subsequent privileged shell activity; alert on child processes spawned by the tenant-upload service, and on shell-history truncation or gaps on the controller host. Because the actor reverted the admin password, an unexplained password-change-then-revert pair in admin account auditing is itself worth investigating.

Hardening. Upgrade to a fixed train — 20.9.9.2, 20.12.7.2, 20.15.4.5, 20.15.5.3, 20.18.3.1, or 26.1.1.2 — as there is no workaround. Restrict which operators hold privileged CLI roles, place the management/northbound interfaces behind a source-IP ACL rather than exposing them broadly, enforce MFA on all administrator accounts, and rotate SD-WAN admin credentials (including the default vmanage-admin) on any controller that may have been exposed before patching. Cisco's Catalyst SD-WAN Hardening Guide carries the vendor's own configuration baseline.

Mandiant publishes the forensic reconstruction behind Cisco SD-WAN Manager CVE-2026-20245

UPDATE · originally covered CVE-2026-20245 — Cisco Catalyst SD-WAN Manager: actively-exploited command-injection to root (no patch) (2026-06-06)

When we first noted CVE-2026-20245 it was a fresh Cisco advisory for a command-injection-to-root flaw in Catalyst SD-WAN Manager with confirmed exploitation but little public detail. Mandiant/GTIG has now published the forensic reconstruction, confirming the flaw was used as a zero-day at a communications service provider from late 2025 through March 2026 — months before the patch (Mandiant/GTIG, 2026-06-24).

The new substance is the kill chain: a peering-authentication-bypass foothold (CVE-2026-20127 / CVE-2026-20182) into SSH as vmanage-admin, then a crafted tenant CSV through the request tenant-upload CLI handler injecting commands that planted a backdoor troot UID-0 account, with anti-forensic clean-up (admin-password change-then-revert, history/syslog deletion). Mandiant names no threat actor. Full mechanics, ATT&CK mapping and host-level detection are in §5.

8x8 confirms Klue/Icarus Salesforce exfiltration in an SEC 8-K Item 1.05 filing

UPDATE · originally covered Klue OAuth-token breach — victim list grows, CRM-API abuse chain detailed (2026-06-21)

US cloud-communications provider 8x8 (NASDAQ: EGHT) filed a Form 8-K Item 1.05 on 2026-06-23 disclosing that an unauthorised party accessed its Salesforce environment on 2026-06-11/12 via a third-party integration — the Klue competitive-intelligence platform — the OAuth-integration vector behind the Icarus extortion campaign already tracked in prior briefs (SEC EDGAR — 8x8 Form 8-K, 2026-06-23).

The filing states the accessed data is limited to contract information, internal sales notes and business contact data (names, business emails, phone numbers, mailing addresses). As a publicly-listed company's mandatory material-incident disclosure, it is the formal confirmation that 8x8 is a named Klue-integration victim, extending the campaign's confirmed-victim list.

Defender takeaway for anyone running SaaS-to-Salesforce OAuth integrations (including EU public-sector users of competitive-intel tooling): audit Connected Apps in Salesforce Setup → App Manager for unexpected or stale OAuth grants, scope connected-app permissions to least privilege, and monitor EventType=OAuthToken in Salesforce Event Monitoring for anomalous token use (T1078.004 Valid Accounts: Cloud, T1550.001 token abuse).

FortiBleed — first full tool-chain disclosure (FortigateSniffer, SNIFTRAN, GPU cracking cluster); Fortinet confirms no new CVE

UPDATE · originally covered FortiBleed — 73,932 internet-facing FortiGate devices exposed, Russian-speaking group cracking credentials into Active Directory (2026-06-18)

New analysis published 2026-06-22 gives the first complete tool-chain picture of the FortiBleed credential-harvesting campaign. The operators deploy a purpose-built Golang tool, FortigateSniffer, that abuses FortiOS's native diagnose sniffer packet diagnostic command to capture authentication traffic on a compromised FortiGate; a second tool, SNIFTRAN, converts the captured traffic to PCAP, which a Python toolkit then parses for cleartext credentials, NTLM hashes, Kerberos tickets and LDAP/SQL auth material across ~24 protocols (BleepingComputer, 2026-06-22; SOCRadar, 2026-06-16).

Fortinet's PSIRT response confirms the campaign uses no new vulnerability — it reuses credentials from the previously-disclosed CVE-2026-24858, CVE-2025-59718 and CVE-2025-59719 plus brute force against devices lacking strong passwords and MFA (Fortinet PSIRT, 2026-06-19; SecurityWeek, 2026-06-22). Reported tradecraft includes a distributed 36-GPU cluster — rented from a generative-AI provider, per BleepingComputer — for offline cracking of the harvested hashes; SOCRadar characterises the operators as Russian-speaking (SOCRadar, 2026-06-16).

The delta for defenders is a concrete detection surface that earlier coverage lacked: FortiOS audit-logs diagnose sniffer packet execution, so hunt for unexpected CLI sniffer invocations and stray PCAP files on the appliance, and — because harvested AD credentials are the downstream prize — treat all domain credentials on any FortiBleed-corpus device as compromised and force a domain-wide rotation, watching for anomalous Kerberos service-ticket requests (event 4769) and new-source Logon Type 3 events (4624) against privileged accounts. Upgrade to firmware with PBKDF2 password hashing to make offline cracking expensive, terminate active sessions, enable MFA and disable external management access.

"Squidbleed" — a 29-year-old heap over-read in Squid's FTP gateway leaks other users' cleartext HTTP credentials (CVE-2026-47729)

Researchers at Calif.io disclosed CVE-2026-47729, nicknamed Squidbleed: a heap buffer over-read in the Squid proxy's FTP-over-HTTP gateway (src/FtpGateway.cc) introduced by a 1997 code commit (Calif.io, 2026-06-18; The Hacker News, 2026-06-22). The root cause is a whitespace-skipping loop that calls strchr(w_space, *copyFrom) without first checking for the string terminator: strchr returns a non-NULL pointer when the search character is the embedded \0, so the parser walks past the end of the FTP directory-listing buffer into adjacent heap memory containing other users' cached HTTP requests. An attacker who controls an FTP server and can induce the proxy to fetch from it (FTP support and TCP/21 are in Squid's default Safe_ports ACL) can leak Authorization headers, session cookies, API keys and other cleartext request content from concurrent users sharing the same proxy worker (SecurityWeek, 2026-06-22). HTTPS relayed via CONNECT tunnels is not exposed; only cleartext HTTP and TLS-terminating proxy setups are. SUSE rates it moderate (CVSS 6.5) and there is no confirmed in-the-wild exploitation. The fixed-version picture is disputed upstream: the patch was merged in spring 2026, but the Squid maintainer first attributed the fix to 7.6 (released 8 June 2026) then corrected that to 7.7, while Debian's assessment is that the referenced commit is already present in 7.6, and SecurityWeek reports the fix shipped in 7.6 (The Hacker News, 2026-06-22; SecurityWeek, 2026-06-22). The safe reading for defenders is to treat the fixed version as uncertain and verify against your own build rather than assuming a single release line is clean Calif.io credits an AI model (Anthropic's "Claude Mythos") with surfacing the strchr edge case during AI-assisted fuzzing — another data point in the AI-assisted-vulnerability-discovery pattern the W25 weekly tracked.

Why it matters to us: Squid is widely deployed as a forward / caching / web-filtering proxy across EU public-sector networks, university perimeters and ISP infrastructure — exactly the multi-user environments where the cross-user leak has impact. Interim mitigation that does not depend on resolving the fixed-version dispute: disable FTP proxying (acl ftp proto FTP + http_access deny ftp, or drop FTP from Safe_ports) where it is not needed, and restrict who can reach the proxy from untrusted/multi-tenant segments. Confirm the fix is present in your actual build (RHEL/Debian/Ubuntu ship 4.x–6.x — check for a backport) rather than trusting a version number. Detection: monitor Squid access.log for ftp://-scheme requests from unusual clients and for worker heap-corruption / crash signals (T1190 Exploit Public-Facing Application; effective outcome resembles T1040 Network Sniffing).

AryStinger: a reconnaissance-and-proxy botnet built on end-of-life D-Link routers and QNAP NAS

QiAnXin XLab disclosed AryStinger, a previously-undocumented botnet its telemetry first observed on 2026-03-12, with English-language follow-up reporting on 2026-06-21 (QiAnXin XLab, 2026-06-17; BleepingComputer, 2026-06-21). Unlike the DDoS- and cryptomining-oriented router botnets that dominate this device class, AryStinger's design centre is pre-intrusion reconnaissance and traffic laundering: infected nodes are enrolled as "Executors" and handed distributed scanning and DNS-brute-force tasks by a C2 controller, and they relay the operator's attack traffic so its true origin is hidden. XLab counts at least 4,300 infected nodes and rising, distributed South Korea 48.5%, China 31.8%, Sweden 6.4%, Malaysia 3.5%, Singapore 2.5%; detection rate on public multi-engine scanning was zero at disclosure.

Initial access — three public CVEs across two device classes. The router variant spreads through CVE-2013-3307 (command injection in Linksys/D-Link models built on the Realtek RTL819X SoC family) and CVE-2016-5681 (a stack-based buffer overflow in the D-Link DIR-850L HTTP service) — both unauthenticated RCE on devices manufactured 2012–2015. From 2026-04-26 a second, NAS-targeting variant began exploiting CVE-2025-11837, a code-injection flaw in QNAP's Malware Remover utility (fixed in build 6.6.8.20251023; QNAP's advisory scopes the affected product to the 6.6.x line — update to the latest build). Mapped to T1190 Exploit Public-Facing Application (T1190). The most infected models — D-Link DIR-850L (75% of nodes) and DIR-818LW (13%) — are end-of-life with no firmware fix (D-Link support bulletin SAP10503), so for the router population there is no patch and replacement is the only remediation.

Post-exploitation and persistence. After exploitation a downloader pulls the current payload from C2, the bot authenticates with a unique Executor ID, and a Dropbear SSH server is deployed on a fixed non-standard port with an iptables rule added to allow inbound C2 traffic — establishing persistent, system-level remote access (QiAnXin XLab, 2026-06-17). This combines T1133 External Remote Services (T1133) for the SSH backdoor with T1562.004 Impair Defenses: Disable or Modify System Firewall (T1562.004) for the firewall change. The router binary masquerades under a system-daemon-like process name (T1036 Masquerading, T1036).

Two malware variants, different capability tiers. The constrained RTL819X C variant carries massdns-style distributed DNS reconnaissance and a NAT-traversal tunnelling module (T1572 Protocol Tunneling, T1572; T1090.002 external proxy, T1090.002). The Go "Standard" variant for more-capable hosts (NAS) bundles off-the-shelf offensive tooling — fscan, ksubdomain, httpx, tlsx — for network-service discovery and subdomain enumeration (T1046 Network Service Discovery, T1046; T1595 Active Scanning, T1595), plus remote command execution and source-level payload execution in Go/Java/Python. C2 is HTTP/HTTPS with Protobuf message bodies under XOR obfuscation; a hardcoded key string embeds a 2024 marker, suggesting the operation predates the 2026 first-sighting.

Why this matters to a Swiss/EU public-sector SOC. The direct exposure is indirect but real: EoL D-Link SOHO routers persist in branch offices, municipal sites, and home-office setups, and QNAP NAS appliances are widely used as departmental file shares — both populations sit on the audience's attack surface, and Sweden's 6.4% share shows European devices are already being conscripted. A node's job is to scan and proxy, so a compromised device inside or adjacent to an organisation's network becomes a launch point for credential brute-forcing and lateral reconnaissance that looks like it originates from trusted infrastructure.

Hunt and detection concepts (no IOCs). On Linux/MIPS network appliances, hunt for an unexpected Dropbear (or any) SSH daemon listening on a non-standard port and for iptables rules added outside change management. On QNAP and other Linux NAS, alert on curl/python (or other interpreters) spawned from the security-utility process tree (T1059.006, T1059.006) and on file writes into /tmp/bin/ by a service account that should not be writing executables. Network-side, watch for bursts of outbound DNS queries consistent with mass subdomain brute-forcing from edge/IoT VLAN segments, and for long-lived outbound SSH from device-management ranges. Inventory edge devices for the affected D-Link models and for QNAP Malware Remover build numbers.

Hardening. Replace EoL D-Link DIR-850L / DIR-818LW (and same-era RTL819X models) — there is no firmware path. Patch QNAP Malware Remover to 6.6.8.20251023 or later. Restrict inbound SSH on management VLANs to known jump hosts, and apply egress filtering so SOHO/IoT segments cannot freely initiate outbound SSH or high-volume DNS. Attribution: XLab claims none; the brief reports the activity as XLab characterises it, not as a named actor.

Brazil's national Cell Broadcast alert platform hijacked to push fake "Extreme Alert" messages to ~30M phones

An unidentified actor gained unauthorised access to Brazil's national Cell Broadcast emergency-alert platform overnight 19–20 June 2026 and sent at least ten unauthorised "Extreme Alert" notifications — the highest-severity tier, reserved for imminent-danger events — to roughly 30 million phones across seven states (The Next Web, 2026-06-20). The Ministry of Integration and Regional Development took the platform offline at 01:30 on 20 June after confirming the intrusion; Brazil's Federal Police opened an investigation and no actor has been formally attributed (a person who claimed responsibility on X had their posts removed, but police have not confirmed the claim). The specific access vector — compromised administrative credential, API key, or platform vulnerability — has not been disclosed. Cell Broadcast is architecturally designed to bypass user opt-outs and to activate devices that are on silent, which is exactly what makes administrative-plane control of it so consequential. [SINGLE-SOURCE] on the primary technical detail

Why it matters to us: This is a demonstrator for a risk class, not a Brazil-specific story. The EU Electronic Communications Code (Directive 2018/1972) mandates Cell Broadcast-based public-warning systems across member states, and Switzerland's Federal Office for Civil Protection (BABS) runs the same technology as ALERTSWISS. The incident points at the administration interface — privileged access to the broadcast console — rather than radio-side spoofing, so operators should prioritise MFA and PAM on alert-platform admin accounts, least-privilege on broadcast-issuing roles, and anomaly detection on outbound broadcast commands (volume, severity tier, off-hours issuance). A false high-severity alert is both a public-safety and a public-trust event.

CVE-2026-50751 — Check Point Security Gateway IKEv1 VPN authentication bypass: public PoC, Qilin affiliate use

Status update on the W24 § 1 item: NCSC-NL updated its advisory on 2026-06-16 to note public proof-of-concept code is now available for the IKEv1 VPN authentication bypass, which a Qilin ransomware affiliate has used for initial access (Help Net Security; NCSC-NL NCSC-2026-0179; daily 06-17). A Remote Access VPN gateway still running the deprecated IKEv1 path is an active ransomware entry point. Apply the Check Point hotfix and disable IKEv1 where IKEv2 can replace it.

CVE-2026-20262 — Cisco Catalyst SD-WAN Manager: authenticated arbitrary file write to root, exploited as a zero-day (CISA KEV)

A path-traversal weakness in the web UI of Cisco Catalyst SD-WAN Manager (formerly vManage) lets an authenticated remote attacker create or overwrite any file on the underlying OS and escalate to root code execution; Cisco patched it after zero-day exploitation and CISA added it to KEV (Cisco PSIRT; daily 06-16). SD-WAN Manager is the centralised control plane for an entire SD-WAN fabric, so a rooted controller is a fabric-wide compromise. Patch on emergency cadence and restrict management-plane access to a dedicated administrative network.

FortiBleed — Russian-speaking operator cracking 86,644 FortiGate credentials into Active Directory

If you did nothing this week: any internet-facing FortiGate whose admin or SSL VPN credentials are in the "FortiBleed" corpus is a live initial-access foothold right now — patch level is irrelevant, because the leaked credential is the weapon, and the operator is already pivoting from validated VPN logins into internal Active Directory.

The FortiBleed dataset surfaced on 2026-06-17 as 73,932 unique FortiGate management URLs (~75,000 devices across 194 countries) paired with valid VPN and administrative credentials (BleepingComputer, 2026-06-17; daily 06-18). By 2026-06-19 the verified count had grown to 86,644 confirmed working credentials and CISA had issued an emergency hardening advisory (SecurityWeek, 2026-06-19; daily 06-20). Fortinet's PSIRT confirmed the campaign ties to previously disclosed incidents (FG-IR-26-060 / FG-IR-25-647) and that the credentials originated from exported device configurations — its position is that this is not a new CVE, the corpus being a reshare of prior-incident data combined with large-scale brute-forcing (Fortinet PSIRT, 2026-06-19) — but that distinction is cold comfort operationally: the credentials validate. The methodology that emerged this week is the load-bearing detail. A Russian-speaking actor intercepts SSL VPN authentication, cracks the captured hashes on a 45-GPU Hashtopolis cluster, and then uses the recovered service and admin accounts to move laterally into internal Active Directory (T1078 valid accounts following T1110 credential cracking).

The escalation that makes this § 1 rather than a routine credential-leak note is the AD pivot plus CISA's mandated response: terminate all SSL VPN sessions, reset every credential, migrate admin-hash storage from the older MD5-crypt scheme to PBKDF2, and enforce phishing-resistant MFA on all remote access. FortiGate is ubiquitous on Swiss and EU public-sector and telco perimeters, so treat any exposed device's local admin and VPN secrets as potentially in the corpus regardless of firmware version. Hunt for sequential VPN authentication failures from rotating residential IP ranges followed by a success and immediate internal RDP/SMB/LDAP reconnaissance, and cross-reference SSL VPN session logs against the Shadowserver notification feed.

FortiBleed reaches 86,644 compromised FortiGate devices; CISA issues emergency hardening guidance

UPDATE · originally covered FortiBleed — 73,932 internet-facing FortiGate devices exposed, Russian-speaking group cracking credentials into Active Directory (2026-06-18)

The FortiBleed SSL VPN credential-harvesting campaign has grown from the 73,932 internet-facing FortiGate devices reported on 2026-06-18 to 86,644 confirmed compromised credentials across 194 countries, and CISA has published an emergency hardening advisory (SecurityWeek, 2026-06-19; CISA, 2026-06-18).

The new detail is methodology and impact: a Russian-speaking actor cracked SSL VPN password hashes with a 45-GPU Hashtopolis cluster, after which the actors pivot into internal Active Directory using harvested service and admin accounts (BleepingComputer, 2026-06-19). CISA's guidance mandates immediate SSL VPN session termination, full credential resets, enforcement of PBKDF2 (replacing the older MD5-crypt admin-hash scheme), and phishing-resistant MFA on all remote access. Defenders should cross-reference SSL VPN session logs against the Shadowserver notification feed and hunt for sequential VPN authentication failures from rotating residential IP ranges followed by a success and immediate internal RDP/SMB/LDAP reconnaissance.

CVE-2026-42530 / CVE-2026-42055 — NGINX: HTTP/3 QUIC use-after-free and HTTP/2-proxy heap overflow, out-of-band F5 patches

F5 shipped out-of-band patches on 2026-06-17 for two critical NGINX flaws (NGINX, 2026-06-17; SecurityWeek, 2026-06-18). CVE-2026-42530 (use-after-free, CWE-416, CVSS v4 9.2): a remote unauthenticated attacker sends a crafted HTTP/3 session that reopens a QPACK encoder stream in ngx_http_v3_module, corrupting worker-process memory — a crash by default, code execution where ASLR is disabled or bypassed; affects Open Source 1.31.0–1.31.1. CVE-2026-42055 (heap-based buffer overflow, CWE-122, CVSS v4 9.2): in ngx_http_proxy_v2_module/ngx_http_grpc_module, but only under a non-default configuration triple — proxy_http_version 2 or grpc_pass, ignore_invalid_headers off, and large_client_header_buffers above 2 MB. Fixed in Open Source 1.31.2 (and 1.30.3 stable), NGINX Plus R36 P6 / 37.0.2.1, and Gateway Fabric 2.6.4. Interim mitigation for CVE-2026-42530 is to remove quic from all listen directives (disabling HTTP/3); for CVE-2026-42055, keep ignore_invalid_headers at its default on. Note the scoring split: nginx.org's own advisory rates CVE-2026-42530 "major" and CVE-2026-42055 "medium" (reflecting the latter's non-default-config gating), while SecurityWeek scores both at CVSS v4 9.2; the brief carries the higher third-party score with the vendor's qualifier noted. F5 reports no in-the-wild exploitation.