Hugging Face turns to Zhipu GLM-5.2 to analyze OpenAI hack
Hugging Face, a New York AI startup, turned to China's Zhipu GLM-5.2 model to analyze data from an OpenAI agent hack, after top US models like GPT-5.6 Sol and Claude Fable 5 refused the job due to strict safety guardrails.
Hugging Face's co-founder Clement Delangue pointed out that defenders need more open access to powerful AI for real-world problems.
GLM-5.2 praised as Zhipu raises $4B
Zhipu's GLM-5.2 is getting a lot of buzz in Silicon Valley for its advanced coding skills and fewer restrictions compared to US models.
Its popularity is soaring: the company just raised about $4 billion in Hong Kong and its stock has jumped nearly nine times since January.
Analysts say US should rethink safeguards
With Chinese AI models gaining ground, some analysts say it's time for the US to rethink its strict safeguards, which can actually hold back efforts to fight cyber threats effectively.