Hackers target AI coding tools to steal credentials, report finds
Original title: Hackers Target AI Coding Tools to Steal Credentials as Researchers Uncover 103 AI-Linked Cyber Incidents
Recorded Future counted 103 AI-linked incidents over a year; only 32 involved AI inside malware or offensive operations.
Recorded Future’s Insikt Group reviewed 103 documented AI-linked incidents from July 31, 2025, through July 31, 2026. Attackers compromised AI coding packages and plugins to steal credentials, plant backdoors, and move laterally, including a late-February proxy package and an April package campaign. Shai-Hulud expanded in May to Claude Code, and The Gentlemen reportedly used language models to shape extortion. Fully autonomous attacks remain uncommon.
Unconfirmed: The 103 incidents are not 103 autonomous attacks or confirmed breaches of AI providers; IoCs were not published.
- Thirty-two cases used AI in malware or operations; nine were state-sponsored intelligence collection.
- The other 71 covered attacks on AI infrastructure, themed delivery, deception, and service abuse.
- Fake Claude Code installers used ClickFix lures and PowerShell stages for credential theft.
- Published indicators omit hashes, domains, URLs, IP addresses, and file paths.
Impact: Stolen credentials, API keys, data, and persistent access
| Product | Affected |
|---|---|
| AI proxy package | — |
| Claude Code | — |
- 1.Separate coding agents from sensitive systems and require approval for consequential actions.
- 2.Audit AI dependencies, plugins, and model repositories for backdoors or malicious command injection.
- 3.Monitor exposed developer API keys, corporate OAuth tokens, and stolen AI subscriptions.
- ATT&CK
- T1195.002 Compromise Software Supply ChainT1566 PhishingT1059.001 PowerShellT1528 Steal Application Access Token
- Products
- Claude Code
- Threat actors
- The Gentlemen
- Malware
- Shai-Hulud
- Organizations
- Recorded FutureInsikt Group
- Countries
- Taiwan
From the original article
3 min read at cybersecuritynews.com
Hackers are targeting AI coding assistants, software packages, and developer integrations to steal credentials and gain access to connected systems. A new threat intelligence report examined 103 AI-linked cyber incidents, showing how tools trusted with source code and sensitive data are becoming entry points for attackers. The research covered July 31, 2025, through July 31, […]
By Tushar Subhra Dutta · cybersecuritynews.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.