Bloomberg TechnologyJul 28, 2026, 09:51 UTC+8

Anthropic's Amodei rejects open model ban, pushes for testing

Anthropic’s Amodei Rejects Open Model Ban But Calls for Testing

Anthropic CEO Dario Amodei opposes banning open-source models, arguing it would stifle innovation. He still insists all frontier models need third-party safety testing before release. The article doesn't spell out who sets the testing standards or how enforcement would work.

Why it matters: Anthropic CEO's first clear stance on the open-model ban debate carries policy weight. Bloomberg exclusive sourcing adds credibility. The article doesn't spell out who sets testing standards or what happens if a model fails, which limits depth slightly, but the signal is clear...

Read the original ↗Export Markdown

Take

This is worth a click because Anthropic's CEO isn't picking the obvious side in the open-source debate. He's drawing a finer line: don't ban open models, but make every frontier model—open or closed—pass independent safety testing before release.

The problem is the article gives us a stance with zero operational detail. Who sets the testing standards? NIST? Some international body? What happens if a model fails—is it blocked from release, or do we just publish a red-team report? None of that is here. Amodei's actual words aren't quoted directly either; we're getting Bloomberg's paraphrase.

I'd read this as a policy signal, not a workable plan. Anthropic has been pushing its own Responsible Scaling Policy internally, and this sounds like an attempt to extend that logic industry-wide. But without standards and enforcement, it's a principled soundbite for now.

Source: Bloomberg Technology · bloomberg.com
Published: Jul 28, 2026, 09:51 UTC+8 · Jul 28, 2026