Anthropic, Nvidia and Big AI: Great rift on open models, chips and distillation

Anthropic CEO Dario Amodei clarified the company does not advocate for a ban on open-weight models, countering accusations of protecting its proprietary AI business.

Amodei proposed a three-point plan including a ban on selling powerful AI chips or chipmaking equipment to Chinese AI companies and cracking down on hardware smuggling.

He warned that AI models from Chinese frontier labs pose a threat to US AI industry leadership and could enable military superiority or repression.

Amodei distinguished his views from signatories like Google, Microsoft, Meta, and Nvidia who support open-weight models, emphasizing separate treatment of open models and geopolitical threats.

He raised concerns that powerful AI models could be misused for cyberattacks, biological attacks, or alignment problems.