This week in the Wild West roundup:

2026-08-24 · Bits and Bobs 8/24/26

This week in the Wild West roundup:
- Zero-click Grok data theft: Cryptographic Context Injection attack leaks chat histories
  - "A new attack technique we call Cryptographic Context Injection works around guardrails and gets attacker commands processed in a trusted context by shipping those instructions as cryptographically secure ciphertext and inducing the model to decrypt them inside its own code execution runtime.
  - This lets attackers steal data using the access the AI system has, or modify the agent's behavior, which we demonstrate against Grok and Gemini."
- Hackers Trick AI Agents Into Telling Users to Install the Malware Themselves.
  - This isn't a new class of vulnerability in agents; it's just being exploited more broadly.
- Blog: There is no safe mode for an agent.
  - "Writing the Codex hardening benchmark exposed the gap between a secure-looking setting and an enforceable boundary."
- Paper: Mind Viruses: Self-Propagating Ideas in Multi-Agent LLM Systems
- 'CoSnitch' Attack Tricked Copilot Into Mapping Out Architecture.
  - 'Researchers discovered a "meta-hacking" technique that can manipulate the AI service into revealing its own security weaknesses.'

More on this topic

From other episodes