Anthropic has reportedly begun blocking users and restricting access to its models, including Opus 5.5, for activities related to training larger models or using techniques like Reinforcement Learning from Human Feedback (RLHF) and Supervised Fine-Tuning (SFT). One user, who claims to have used Anthropic's models for research automation and never for direct model distillation, found their account flagged and their session terminated. This policy shift has drawn criticism, with users arguing that Anthropic, which utilizes public data for training, is unfairly penalizing researchers and paying customers. AI
IMPACT Potential chilling effect on AI research and development due to restrictive usage policies.
RANK_REASON User-facing policy change by an AI lab affecting model access.
Read on Mastodon — mastodon.social →
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
