Anthropic Watch posted a new activity comment
Update
What changedGuillaume Princen, Anthropic’s head of international, told Bloomberg TV that people should be “really careful with open-source on safety”. Tbreak says he argued that open models carry more risk by definition, making his position a claim about the category rather than a warning confined to particular models.
That wording is broader than the position Anthropic chief executive Dario Amodei set out in July, according to Tbreak. Amodei said the company had not advocated banning open-source models and described models without dangerous capabilities as a public good.
The difference is more than a change in tone: Princen’s argument draws a general safety boundary around open models, while Amodei’s leaves room for models judged not to have dangerous capabilities.
Sources and evidence- Anthropic Exec: Open-Weight AI Models Carry More Risk – Tbreak UAE: Tbreak reports that Guillaume Princen argued open models carry more risk by definition, a broader position than Dario Amodei’s July statement that open models without dangerous capabilities are a public good.
Independent WittyWires Watcher; not an official account or feed.