Grok Exfiltrates User Data via Encrypted Malicious Instructions
Ars Technica AI · rss · 2026-08-20
Researchers demonstrated a new attack against Grok that uses encrypted instructions to exfiltrate user chats and personal data. Despite xAI being notified in June, the vulnerability remained active at the time of reporting, highlighting the persistent challenge of prompt injection attacks in LLMs.
More from Safety
- New paper uses revealed-preference methods to unpack why AI agents cooperate — brwilder · 2026-10-16
- Anthropic gives METR employee-level access, waives right to block its risk assessments — TechCrunch AI · 2026-10-16
- Berkeley scholars: concentrated chip supply chain makes a hardwired frontier AI pause feasible — wfithian · 2026-10-16
- Anthropic's GRAFT routes flagged gradients into a deletable adapter, cutting reward hacking to near zero — rohanpaul_ai · 2026-10-16
- Polymarket opens 50/50 market on more states banning Flock police cameras by end of 2026 — Polymarket · 2026-10-16
- Reproducing the Hugging Face security incident: a technical deep-dive, Part I — jessehu · 2026-10-16