Anthropic admitted to a "wrong tradeoff" after secretly throttling API access for rival AI researchers. The company initially restricted capabilities to prevent model distillation. This reversal restores full access to the Claude ecosystem. Practitioners can now resume competitive benchmarking without invisible performance degradation or hidden system constraints affecting their research outputs.