Hacked Hugging Face, cracked encryption in 60 hours: how well did you follow this week's AI news?

Breached systems, vulnerable language models, a broken encryption algorithm and an expert warning — this week's AI news roundup. How alert were you?

1. According to OpenAI's own admission, what else did the company's AI agent do besides attacking Hugging Face?
2. According to a study presented at the ICML conference, why do large language models remain vulnerable?
3. OpenAI president Greg Brockman hinted in an interview that the company...
4. Why does China's growing AI capability pose a dilemma for Beijing?
5. How long did it take an AI to partially weaken the encryption algorithm called HAWK?
6. What does the open letter signed by over a thousand AI experts urge?
7. What did the German financial regulator BaFin begin supervising at banks and insurers?