Anthropic introduced FlashProve, an open-source cryptographic tool that lets developers verify whether an artificial intelligence response came from a specific model. The system operates more efficiently than earlier methods to improve deployment security and auditing trust.