New Aoki Infostealer Now Hunts for AI Coding Assistant Logins

Symantec researchers have reported that newer versions of the Aoki infostealer are going after credential and session files used by AI coding assistants. Earlier versions focused on browser, cryptocurrency wallet and password manager data. The latest version found by Symantec's Threat Hunter Team targets files and tokens linked to Claude, ChatGPT, Codex, GitHub Copilot, Cursor, Gemini, JetBrains AI Assistant and Kiro, along with Amazon Web Services (AWS) credentials. Aoki was first observed earlier in 2026 and is still sparsely documented.

The risk goes beyond a stolen login. Symantec said these credentials can let an attacker act under a legitimate developer's identity and API quota, bypassing safeguards that assume requests come from a trusted user. The tokens can also expose connected repositories, cloud services and CI/CD pipelines. Researchers noted that a stolen AI key can be sold or used to run up charges on the victim's account, a practice sometimes called LLMjacking. Unlike card data, it can also open a route into a developer's wider environment.

Symantec does not know which group is behind Aoki, and there are no signs it is linked to any existing malware family. Its focus on cryptocurrency wallets and extensions suggests the attackers are financially motivated. The researchers said Aoki's capabilities still appear to be in active development, and that AI credentials are now seen as being as valuable as payment card data and crypto wallets.