DeepMind CEO calls for an independent standards body to regulate frontier AI

The CEO of DeepMind proposed creating an independent standards body to regulate frontier AI models before their release. The proposed organization would be modeled after a financial industry regulator, with labs voluntarily submitting models for review up to 30 days before release, and eventually making such review mandatory for deployment in the US market. This proposal addresses criticism of previous government reviews of AI models, which were criticized for lacking technical expertise and transparency in decision-making. The standards body would be backed by the U.S. government but funded and operated independently by the AI industry, staffed with technical experts and open source representatives, and could help address concerns about creating a formal government AI regulator.