// AI threat desk · 18 Oct 2026
Home/AI-ATK · AI-Powered Attacks
AI-Powered AttacksMediumAICVSS not assigned

Fake AI campaigns linked to 1,258 analyses by ANY.RUN

ANY.RUN linked 1,258 public sandbox analyses to fake AI campaigns targeting businesses between April and October 2026.

· 1 min read
AI-generated image of a security analyst reviewing blurred sandbox results and suspicious files on multiple monitors.
AI-generated image, not a photo of the event.
Key takeaways
  • ANY.RUN linked 1,258 public sandbox analyses to fake AI campaigns between April and October 2026.
  • ANY.RUN said the campaigns encourage employees to execute malicious commands, install compromised applications, or submit credentials through fraudulent authentication pages.
  • The Hacker News said the lures impersonate AI services and appear in sponsored search results, legitimate hosting services, shared AI conversations, and counterfeit download websites.

ANY.RUN linked 1,258 public sandbox analyses to fake AI campaigns targeting businesses between April and October 2026.

The Hacker News said attackers impersonate Claude, ChatGPT, GitHub Copilot, Gemini, and other AI services. It said the lures appear in sponsored search results, legitimate hosting services, shared AI conversations, and counterfeit download websites.

On this page

What happened

ANY.RUN researchers also identified scheduled tasks that launch browsers to capture sessions, along with encoded download instructions and suspicious PowerShell installation commands.

US businesses made up 16% of country-attributed targets

The Hacker News said the campaigns also increase operational pressure on MSSPs serving US businesses, which accounted for 16% of country-attributed targets.

A managed security service is a network security service outsourced to a service provider, according to Wikipedia.

ANY.RUN reported fewer escalations with its sandbox

ANY.RUN reported 30% fewer Tier 1 to Tier 2 escalations and a median threat detection time of 14 seconds when using its sandbox.

It said its Threat Intelligence Feeds provide 99% unique indicators of compromise from the latest malware and phishing, while 30% of unique malicious indicators discovered by customers during sandbox investigations are already available in those feeds.

FAQ

What happened in the fake AI campaigns?

ANY.RUN linked 1,258 public sandbox analyses to fake AI campaigns targeting businesses between April and October 2026.

Which AI products did the campaigns impersonate?

The Hacker News said attackers impersonate Claude, ChatGPT, GitHub Copilot, Gemini, and other AI services.

How do the fake AI campaigns reach employees?

The Hacker News said the lures appear in sponsored search results, legitimate hosting services, shared AI conversations, and counterfeit download websites.

What can the fake AI pages make employees do?

ANY.RUN said the campaigns encourage employees to execute malicious commands, install compromised applications, or submit credentials through fraudulent authentication pages.

Sources

  1. Fake AI Attacks Spread Across US Companies: How MSSPs Can Stop Incidents Early, The Hacker News
  2. Hackers Use Google Ads and Bing to Spread Fake Claude Installers, Windows Report
  3. Managed security service, Wikipedia
How we checked this story
ClaimSourceStatus
The Hacker News said attackers are impersonating popular AI products to distribute malware, steal browser sessions, and compromise business accounts.The Hacker NewsAttributed
ANY.RUN linked 1,258 public sandbox analyses to fake AI campaigns between April and October of 2026.ANY.RUN, via The Hacker NewsAttributed
ANY.RUN said the campaigns use familiar AI brands to encourage employees to execute malicious commands, install compromised applications, or submit credentials through fraudulent authentication pages.ANY.RUN, via The Hacker NewsAttributed
The Hacker News said attackers impersonate Claude, ChatGPT, GitHub Copilot, Gemini, and other AI services.The Hacker NewsAttributed
The Hacker News said the lures appear in sponsored search results, legitimate hosting services, shared AI conversations, and counterfeit download websites.The Hacker NewsAttributed
The Hacker News said the command launches mshta.exe, retrieves a malicious HTA file, and triggers an encoded PowerShell execution chain.The Hacker NewsAttributed
ANY.RUN reported 30% fewer Tier 1 to Tier 2 escalations and a median threat detection time of 14 seconds when using its sandbox.ANY.RUN, via The Hacker NewsAttributed
ANY.RUN said its Threat Intelligence Feeds provide 99% unique indicators of compromise from the latest malware and phishing.ANY.RUN, via The Hacker NewsAttributed
30% of unique malicious indicators discovered by ANY.RUN customers during sandbox investigations are already available in its Threat Intelligence Feeds.ANY.RUN, via The Hacker NewsAttributed
ANY.RUN researchers identified suspicious PowerShell installation commands, mshta.exe execution, encoded download instructions, and scheduled tasks launching browsers for session theft.ANY.RUN, via The Hacker NewsAttributed
Windows Report said hackers use Bing redirects in Google ads to spread fake Claude installers that trick macOS users into running malicious commands.Windows ReportAttributed
A managed security service is a network security service outsourced to a service provider.WikipediaConfirmed

Could not verify

  • Whether the campaigns were exploited through a software vulnerability is not established
  • How many organizations or users were affected is not established
  • Whether the campaigns described by S1 and S3 are linked is not established
  • Whether the reported performance figures were independently verified is not established
See all in AI-Powered Attacks →
Explore with AI
ChatGPT
Claude
Perplexity
Grok