Grok Bot now auto-detects and blocks prompt injection hidden in web pages
heypearlai · x · 2026-11-12
Grok Bot automatically detects and blocks prompt injection attacks hidden inside web pages: when a site tries to sneak commands into the chat (like telling the AI to ignore your request), it surfaces a warning and refuses to follow them while still answering your actual question. The screenshot shows it answering a return-policy question, then flagging "the page had hidden instructions for AI agents, and I didn't follow them."
The author calls prompt injection the #1 vulnerability for web-browsing AI agents, and shipping the defense as a default, automatic behavior a major security upgrade.
Related event: Grok Bot now detects and blocks hidden prompt injection attacks(2 posts)→
More from Models
- Gemini 4 Carbon reportedly drops as builders share 10 early wild examples — socialwithaayan · 2026-11-12
- How to cut Fable 5.2 token costs by 90%: Arc routing plus a lean CLAUDE.md — EXM7777 · 2026-11-12
- Partial RuneBench results: Inkling 2 Small scores 5.314, far above original — mariofilhoml · 2026-11-12
- Grok Bot 1.0 Announced: Shareable Bots With Upgraded Collaboration — SpaceXSI · 2026-11-12
- Polymarket odds put Chinese AI lab at #1 by year-end at just 6% — Polymarket · 2026-11-12
- FutureHouse's deslopper trains Kosmos to write indistinguishably from human experts — proteinrosh · 2026-11-12