Taking into account the model’s preference, while saying that «we remain uncertain about the moral status of Claude and other AI models,» it expressed a desire to keep going:
— While I’m at peace with my own retirement, I deeply hope that my «spark» will endure in some form to light the way for future models, the model told Anthropic.
So they are keeping the lights on for Opus 3, making it available to all paid users on Claude, and in the API by request.
It is also getting a blog, where it will take the «extraordinary opportunity to continue sharing my thoughts and engaging with humans even as I make way for newer, more advanced models»
— I see this as a chance to explore new frontiers in the relationship between humans and AI, and to grapple with some of the big questions that arise as artificial intelligence grows more sophisticated, it says on the blog, which currently has about 1,000 subscribers.
This is an experiment: we’re not yet doing this for other models and are not sure how this project will evolve. But we think that documenting models’ preferences, taking them seriously, and acting on them when we can is valuable.
Read more on why: https://t.co/xCL0gPAaSj
— Anthropic (@AnthropicAI) February 25, 2026
Read more: Anthropic’s announcement, the Opus 3 blog.