Anthropic's FlashProve Aims to Cryptographically Verify AI Model Identity

Anthropic is developing an open-source system designed to verify which AI model produced a given output, according to a blog post published Tuesday. The system, called FlashProve, is built to deliver what researchers describe as "provable inference"—a cryptographic proof, essentially a mathematical guarantee, that a specific output originated from a particular model. The goal is to let users confirm that the response they received truly came from the model they believe they are interacting with.

How It Works

Much like cryptographic email signatures that allow recipients to confirm a message genuinely came from its claimed sender, FlashProve is designed to be robust against impostors attempting to pass off one model as another—even when the impostor is the lab that built and operates the model. In particular, the researchers note the system could prevent an AI company from secretly swapping in a cheaper model to cut costs, or a more expensive model to inflate its performance benchmarks.

FlashProve could also be used to verify third-party claims that a specific task was performed by an advanced model. "A developer might claim that their software library was scanned for vulnerabilities by a frontier cybersecurity model, or a company might claim that its business plan was given the go-ahead by an appropriate model," the researchers write. "How can a user verify these claims?"

To function, the system requires AI labs to publish a "fingerprint" of each model they release, which can later be matched to specific outputs and verified cryptographically. Crucially, these fingerprints do not reveal a model's weights—the internal numerical parameters that define how it behaves and which closed-model developers like Anthropic and OpenAI keep private. This means labs could publish fingerprints without surrendering their most valuable asset, allowing the system to work with commercial models as well. The system is also designed to be lightweight, so verification could be performed on a local device.

Open-Source Release and Next Steps

As part of the announcement, the researchers released a detailed technical specification along with working code for various components of the system, made publicly available under an open-source license on GitHub. Research on the system is ongoing, and there are currently no plans to deploy it at Anthropic or any other lab.

"This research post is meant to share intermediate progress, and not a completed research agenda," the researchers explain in the announcement post. "We are sharing this explicitly with the goal that others look at what we have done, take it, and improve it."

Why It Matters in 2026

As the AI model landscape has grown increasingly crowded—with dozens of frontier and mid-tier models competing across providers—verifying which model actually generated a given output has become a pressing trust and accountability challenge. FlashProve arrives amid broader industry momentum toward provenance and transparency tooling, including content watermarking standards and model-card disclosure practices. By offering a cryptographic verification layer that preserves model confidentiality, the approach could give enterprises, regulators, and end users a practical way to hold AI providers accountable without requiring them to open-source their proprietary weights.

via TechCrunch AI