Anthropic develops FlashProve to verify which AI model produced a response

Anthropic is building a system called FlashProve that would let people confirm which AI model produced a given response. The company's work was first reported on 4 November 2026.

The tool is designed to check the origin of an AI answer. A user could use it to see whether a response truly came from the model that claims to have written it. The sources describe the aim in these terms, and do not give further technical detail on how it works.

Open-source release

A later report said Anthropic has introduced FlashProve as an open-source tool, meaning its code is available for others to inspect and use. That report, published at 09:10 IST the same day, described the tool as a way to verify AI models. The earlier report described the system as still in development.

The sources do not say how widely the tool is available or when it can be used in everyday products. They also do not give any figures on accuracy or performance.

Why identity matters

AI systems from different companies can produce text that looks alike, which makes it hard for a reader to know where an answer came from. A verification tool would aim to give people a way to check that claim. The sources do not say whether Anthropic has tied the tool to any particular product or service.

For readers in India, where many people already use AI chatbots for work, study and daily questions, a tool of this kind could help them judge whether a reply came from the model they chose. The sources do not say whether the tool supports any language or service used in India.

The reports do not say what Anthropic plans to do next, or whether other companies will adopt the tool. Those details are not yet known.