Anthropic CEO Dario Amodei has defended the company’s position on open-weight AI models after criticism from Palantir CEO Alex Karp. Amodei said Anthropic has never argued for a ban on open-weight models, pushing back on the idea that the startup wants tighter access rules mainly to protect its own market position.
The dispute centers on a bigger argument inside the AI industry: whether broad access to powerful models helps the people trying to defend against threats more than it helps bad actors. Amodei challenged that assumption, suggesting the benefits are not as one-sided as some supporters of open-weight releases claim.
His response also appears aimed at separating Anthropic’s safety-focused messaging from accusations that it is trying to limit competitors. That issue has become more sensitive as AI companies face scrutiny over how they talk about openness, security and commercial advantage at the same time.
The exchange highlights how divided major tech and defense-linked firms remain on AI access. As companies race to build and deploy more capable systems, the debate over open-weight models is increasingly becoming a fight over security, competition and who should control the most advanced tools.
