Tip
Anthropic CEO: No blanket ban on open-source models
Anthropic CEO Dario Amodei published an official stance, clearly opposing a blanket ban on open-weight models. The core concern is that authoritarian governments could leverage models to enhance military and surveillance capabilities, and that open-source models' safety protections can be removed and weights cannot be recalled. Anthropic advocates for stricter controls on exports of advanced chips and cracking down on industrial-scale knowledge distillation; it also requires that all sufficiently powerful models (whether open-source or not) must pass mandatory testing on cyber, biological, and alignment risks before release.
Read the original (opens in a new tab)
News stream data aggregated by AI HOT