A Windows data thief called Aokigahara copies login files for AI apps like ChatGPT Desktop, Claude and Cursor, and a password reset alone may not lock thieves out.
Security firm Check Point has described a new Windows data-stealing program called Aokigahara. It hunts for the files that keep you signed in to AI apps, including ChatGPT Desktop, Claude, Codex and Cursor. It also goes after saved developer and cloud logins. Check Point published its findings in an October 23 report.
Aokigahara is an infostealer. That is a type of malware that quietly copies sensitive files from an infected computer and sends them to the attacker. This one has a module that looks for AI app data. It also checks Windows Credential Manager, the built-in place where Windows stores saved logins.
A second module goes after Git logins, SSH files that may hold private keys, and settings for AWS, Azure and Google Cloud. Git is a tool developers use to manage code. Not every file it targets necessarily holds a password or token.
Many apps keep you signed in using a session token, a small file that proves you already logged in. If a thief copies a working one, they may get into your account without your password. Check Point says that could expose private chats, uploaded material or development information. Attackers might also use paid services or resell accounts. These are possible outcomes, not confirmed activity by this group.
That is why changing a password alone may not be enough. Stolen sessions can stay usable. Check Point also raised the chance of data extortion if stolen access reveals sensitive business content.
This is not a breach of ChatGPT, Claude, Codex or Cursor. The report describes theft from computers that are already infected, not a flaw in those services. The material does not say how the malware gets onto a computer. Check Point also says it has not confirmed how the operators make money from what they steal.
Windows users of these AI tools should take note. Developers and businesses with cloud accounts face the most risk, since the malware also looks for code and cloud access. The steps in the alert box above cover what to do if a computer may be infected.
Check Point says an earlier version may have gone by the names Kamidere or Stygian, and it has published a detection rule for security teams. Businesses should treat AI app sessions like any other stolen login. That means checking recent activity in AI, code and cloud accounts for anything unfamiliar.
If a Windows computer with AI tools may be infected, sign out of every session and replace keys from a clean device, because a password reset alone may not be enough.
A user says Anthropic's Claude Code put his email in web requests, then scam messages arrived. The link is unproven, but it shows what AI tools can share.
Read more: User says Claude Code sent his email address to websitesA viral ChatGPT prompt turns what the chatbot remembers about you into a room image. Posting it publicly could give scammers clues to target you.
Read more: Viral ChatGPT room image could hand scammers clues about youEA has re-released 25 classic PC games at a discount until November 19, and Bethesda has finally given The Elder Scrolls VI a subtitle: Sentinel.
Read more: EA brings 25 classic PC games back, and Elder Scrolls VI gets a name
