Microsoft Threat Intelligence: AI-brand impersonation drives Lumma Stealer and Vidar delivery via signed binaries
Microsoft Threat Intelligence documents a campaign by Storm-3075 (initial-access broker) and Fox Tempest (malware-signing-as-a-service operator) that weaponises public enthusiasm for AI tools, impersonating ChatGPT, Claude, DeepSeek and Microsoft Copilot through SEO poisoning, malvertising and multi-stage redirection chains (Rebrandly → CAPTCHA gate → credential-harvesting landing) (Microsoft, 2026-06-08). Downloaded binaries are code-signed with certificates obtained through Fox Tempest's MSaaS operation (T1553.002), suppressing initial detection; payloads include Lumma Stealer, Vidar, Hijack Loader and Oyster, with fraudulent GitHub repositories used for payload staging. Microsoft's separate analysis details the Fox Tempest malware-signing-as-a-service operation that supplies the certificates (Microsoft, 2026-05-19).
Why it matters to us: Code-signing is no longer a trust anchor here — a valid Authenticode signature on a fresh "AI tool" installer is consistent with this chain. Detection concepts: Sysmon EID 1 for browser-parented processes spawning infostealer-family command lines; EDR process-injection alerts for Hijack Loader. Phish-resistant MFA (FIDO2/passkeys) removes the downstream AiTM credential-replay value even when an endpoint is seeded.
ATT&CK mapping
1 technique mapped from the cited reporting · MITRE ATT&CK v19.2
Defense Impairment TA0112
T1553.002Subvert Trust Controls: Code Signing
Adversaries may create, acquire, or steal code signing materials to sign their malware or tools. Code signing provides a level of authenticity on a binary from the developer and a guarantee that the binary has not been tampered with. The certificates used during an operation may be created, acquired, or stolen by the adversary. Unlike Invalid Code Signature, this activity will result in a valid signature.
AI-generated · no human review · this permalink is the shareable record for the finding · verify operationally critical claims against the linked primary source.