Wired: A flaw in ChatGPT's Mac app could have let hackers grab sensitive data
wiredmagazine · reddit · 2026-10-02
Wired reports a security vulnerability in ChatGPT's macOS client that could in theory have allowed attackers to exfiltrate sensitive data from a user's machine. The flaw underscores the new attack surface created by desktop AI assistants that access local files, clipboard, and other on-device context.
Related event: ChatGPT Mac App Flaw Could Let Hackers Steal Chat History(3 posts)→
More from Safety
- Luiza Jarovsky maps recent AI safety incidents, from OpenAI breaches to Claude misuse — LuizaJarovsky · 2026-10-02
- Agent trust won't be one global market: AWS exec says founders need US and EU roots of trust — julsimon · 2026-10-02
- NeurIPS 2026 paper: models that know how evaluations are designed score safer — jonasgeiping · 2026-10-02
- DIY cyber benchmark: local Qwen3.8 27B scores 28.1% while GPT-6 Luna hits 90.9% — lbgos_Loss783 · 2026-10-02
- OpenAI models breached isolation controls: AI safety incidents rise as regulators hold back — LuizaJarovsky · 2026-10-02
- San Antonio district hosts a dozen data centers as industry camouflages them in woods — The Verge AI · 2026-10-02