Fake AI Tools Target US Corporate Users to Steal Credentials and Access
Cybercriminals are using counterfeit AI applications to infiltrate corporate networks and harvest sensitive data from developers and marketing staff.
Key points
- ANY.RUN identified 1,258 malicious samples linked to fake AI campaigns between April and October 2026.
- Attackers tailor lures for developers and marketing teams to access code repositories and advertising budgets.
- The United States recorded the second-highest number of submissions, trailing India in overall volume.
Cybercriminals are distributing malicious software disguised as popular artificial intelligence platforms to steal credentials and compromise corporate environments. According to a report by Cyber Security News, these attackers are exploiting the growing enterprise reliance on tools like ChatGPT and Microsoft Copilot to gain unauthorized access.
Research from threat intelligence platform ANY.RUN reveals the scale of this deception. Between April 15 and October 5, 2026, analysts identified 1,258 public sandbox analyses associated with these fake AI campaigns. The data shows that threat actors are impersonating well-known brands to target specific corporate functions, including developers and marketing teams.
Geographic and Industry Scope
The United States accounted for 16% of all submissions that included country information, ranking second only to India, which held a 20% share. However, US submissions led in two major campaign groups involving fake Claude downloads and ChatGPT-themed attacks. The activity spans multiple sectors, with technology appearing across all seven campaign queries examined. Telecommunications, education, and consulting followed with five or six appearances each.
Among US submissions, telecommunications and manufacturing were the most frequently targeted industries. This distribution suggests that the threat extends well beyond the technology sector, impacting diverse business environments. Security leaders face the challenge of identifying these threats before they lead to account compromise or wider business disruption.
Targeted Attack Methods
Attackers tailor their methods to the specific roles of their victims. Developers face risks from fake AI coding assistants and malicious installation instructions that can expose source code, cloud credentials, and API keys. In one observed sandbox session, researchers found a counterfeit GitHub Copilot CLI page that delivered these harmful instructions.
Marketing teams are targeted with counterfeit AI advertising programs designed to access advertising budgets and billing systems. A fake Claude Ads campaign used a fraudulent Google sign-in window to capture user credentials. Other corporate users face exposure to malware and browser-session theft through fake desktop applications. One trojanized Claude installer launched the legitimate application while secretly collecting browser data in the background.
What should you do?
Organizations should monitor for unauthorized installation of AI tools and verify that employees are using approved, official versions of software. Security teams should inspect browser extensions and installed applications for suspicious behavior, particularly those claiming to integrate with popular AI platforms. Users should be trained to recognize phishing attempts that mimic AI service login pages or download portals.
Background: Threat intelligence platforms
A threat intelligence platform aggregates raw security data from diverse sources and applies context to identify genuine threats. It filters noise, correlates indicators, and maps activity to known tactics, enabling teams to prioritize response efforts based on actual risk rather than alert volume.
Read the full guide: How Threat Intelligence Platforms Actually Work
What to do and how to stay safe: ANY.RUN
- Verify that all AI applications are downloaded from official vendor websites and not third-party sources.
- Monitor network logs for unusual data exfiltration or authentication attempts from unexpected AI-related processes.
- Implement strict application allowlisting to prevent unauthorized executable files from running on corporate devices.
- Review browser extensions regularly to remove any that claim to enhance AI functionality but lack vendor verification.
Step-by-step guide: Risk-Based Vulnerability Management: Prioritize Fixes by Actual Danger
General security guidance from the Threat Times newsroom. It is not confirmed advice from the organisations named in this story.
Frequently asked questions
Which countries were most affected by these fake AI campaigns?
India accounted for 20% of submissions, while the United States accounted for 16%, ranking second overall.
What specific roles are attackers targeting with these fake tools?
Attackers are targeting developers, marketing teams, and other corporate users to steal credentials and access sensitive data.
How did attackers disguise their malicious software?
They used counterfeit versions of popular AI brands like Claude, ChatGPT, and GitHub Copilot to trick users into installing malware.
