Anthropic reverses course on silently throttling rival AI researchers
Anthropic backs down on policy to invisibly degrade Claude Fable 5 performance for AI researchers after backlash.

Anthropic had planned to invisibly degrade Claude Fable 5's performance for users trying to train competing AI models. After sharp backlash from the research community, the company is now backing down. "We made the wrong tradeoff and we apologize for not getting the balance right," Anthropic told WIRED.
Going forward, any protective measures will be visible to users. Dean Ball, a former White House AI advisor, called the covert approach "shockingly hostile." Will Brown from open-source startup Prime Intellect put the core criticism bluntly: "It felt like Anthropic was saying to the public, 'We don't trust anybody else to do AI research. We are the only ones who have to do AI research.'" That's not the only controversy around Fable 5, though.
Claude Fable 5 requires data retention to run its new safety classifiers—prompts and outputs are stored for up to 30 days, and up to two years if policy violations are flagged. That's a dealbreaker for some. Microsoft is restricting the model internally because of it, according to The Verge: while all other Claude models run under zero-data-retention rules, Fable 5 doesn't even show up in the internal model picker for Microsoft's GitHub Copilot version.
Why this matters: Anthropic's initial decision to silently throttle rival AI researchers sparked concerns about the company's commitment to openness and collaboration in the AI community. By reversing course, Anthropic is taking a step towards transparency, but the data retention policy for Claude Fable 5 remains a point of contention. This could have broader implications for the development of AI models, as researchers and businesses may be hesitant to work with models that require data retention.
The decision also raises questions about the balance between protecting intellectual property and promoting innovation in the AI space. As the AI landscape continues to evolve, Anthropic's actions will be closely watched by the research community and industry stakeholders.
Source: The Decoder