Preparing for future AI risks in biology - Scaling security with responsible disclosure
- Disrupting malicious uses of AI: June 2025
- Operation “ScopeCreep”: Russian-speaking malware development
- Operation “VAGue Focus”: Social engineering and influence activity
- Operation “High Five”: Philippines political comments
- Operation “Wrong Number”: AI-assisted task scam
- Vixen and Keyhole Panda: China-linked cyber operations
- Deceptive Employment Scheme: IT worker activity
- Operation “Helgoland Bite”: German-language influence activity
- Operation “Sneer Review”: China-origin influence activity
- Operation “Uncle Spam”: US polarization influence activity
- STORM-2035: Recidivist influence activity
- Addendum to OpenAI o3 and o4-mini system card: OpenAI o3 Operator
- Addendum to o3 and o4-mini system card: Codex
- Expanding on what we missed with sycophancy
- Sycophancy in GPT-4o: what happened and what we’re doing about it
- Thinking with images
- OpenAI o3 and o4-mini System Card
- Moving from intent-based bots to proactive AI agents