Skip to content
AgentProv

2026-07-23 · 2026-Q3

"FakeAgent": Malicious Claude Artifact and Bing Ads Deliver SectopRAT to 29 Organizations

Outside the model: Offensive useAgent tooling was not the attack surface, so no layer applies.

Fixable
Architectural
Vendor response
Unknown
In the wild
Yes
Actor
Not applicable

Target

Corporate users searching Bing for the Claude desktop application; at least 29 organizations compromised between July 21 and July 22, 2026

Impact

A sponsored Bing result pointed at a public Artifact hosted on Anthropic's own claude.ai domain, a shareable page anyone can publish. Because the first hop sat on a legitimate vendor domain, the lure carried unusual trust. The Artifact redirected to attacker infrastructure serving a fake `ClaudeDesktop.exe` installer that dropped SectopRAT, a .NET remote access trojan that harvests saved passwords, browser cookies, autofill data, payment card details, and files, and that uses VMProtect and other anti-analysis techniques. The malicious page accumulated about 7,100 views before it was reported and taken down; investigators found command-and-control data stored on the Ethereum blockchain

Root cause

A user-generated content hosting feature on a trusted AI vendor domain, combined with paid search placement, was used to launder the reputation of a malware download; there was no compromise of Anthropic systems. Documented by Huntress

Sources

Catalog

Cite this: AgentProv incident record: "FakeAgent": Malicious Claude Artifact and Bing Ads Deliver SectopRAT to 29 Organizations (2026-07-23). https://agentprov.com/incidents/2026-07-23-fakeagent-malicious-claude-artifact-and-bing-ads-deliver-sectoprat-to-29