Anthropic has reversed a controversial policy that would have secretly limited researchers’ ability to use its Claude AI model for developing competing AI systems, according to WIRED. The company changed course after researchers publicly criticized the policy.
According to WIRED, the original policy would have covertly restricted Claude’s capabilities when used for AI model development, potentially impacting legitimate research activities. Researchers who spoke out against the measure characterized it as potentially “sabotaging” their work with the AI assistant.
The reversal represents a significant shift in Anthropic’s approach to governing how its AI technology can be used by the research community. The company’s initial policy had raised concerns about transparency and the ability of researchers to freely conduct AI development work using Claude as a tool. By walking back the policy, Anthropic appears to have responded to feedback from the research community about the importance of maintaining open access to AI tools for legitimate academic and scientific purposes.