Anthropic open-sources FlashProve to verify which AI model wrote a reply

AIAnthropic has open-sourced FlashProve, a cryptographic system that lets a server prove which AI model produced a given response. The post says it is far more efficient than prior open-source methods, and that a single Nvidia H100 produced proofs for Meta's Llama 4 Maverick at 110 tokens per second, each about 19MB, which a regular CPU checks in about 17 seconds.

Rohan Paul

@rohanpaul_ai

View on X

Anthropic just open-sourced FlashProve, which lets you check which AI model actually wrote a reply

checkable on an ordinary CPU.

e.g. if you pay for a top AI model through a reseller, you can’t check whether it quietly served a cheaper one. Anthropic’s FlashProve lets the server prove mathematically which model answered.

Anthropic tested it on Meta’s Llama 4 Maverick, a large 400B-parameter open model, where a single Nvidia H100 GPU produced proofs at 110 tokens/second, each about 19MB.

A regular CPU checks each proof in about 17 seconds, while VerInf, the only earlier system known to prove this model, made 93.6GB proofs that took almost a day to check.

Claude built the system, and because its code is hard to audit, the security guarantees are mathematically proved in Lean 4, a proof-checking language,

Anthropic@AnthropicAI

New Anthropic research: progress on provable inference for large language models. Here, we introduce FlashProve. Building on established cryptographic techniques, FlashProve allows developers to verify whether a given response came from a given AI model. It’s substantially more efficient than prior open-source methods for doing so. These proofs have a range of applications. They can, for instance, make it harder for attackers to hijack a chat, or enable safety evaluators to verify that the model they’re testing is the same one being served to customers. For our full write-up, see here: https://www.anthropic.com/research/provable-inference Further progress in provable inference will likely be an important part of improving trust in AI deployment. To that end, we're open-sourcing FlashProve’s code, here: https://github.com/anthropics/FlashProve We’re grateful to our collaborators: Alessandro Chiesa, @GiacomoFenzi, @gwberg95, @miranda__christ, Rahul Dalal (http://rdalal.github.io) and Ben Mayer (@transmissions11).

View quoted post on X

Source details

Rohan Paul · x.comPublished Nov 3, 3:15 PM ET · added here 3:34 PM ET

Original title: Anthropic just open-sourced FlashProve, which lets you check which AI model actually wrote a reply

Anthropic open-sources FlashProve to verify which AI model wrote a reply

Topics