What matters in AI.

Subscribe

Proof puts its agent approval check in ChatGPT

A person can add the Proof app in ChatGPT, and a business can check a signed approval for each task.

Proof is a company that checks identity and approval for online tasks. It released the first version of its Model Context Protocol server in ChatGPT. This standard connects AI assistants to external tools. The release puts x401, an open protocol that Proof launched on 15 October, to ChatGPT users in the app. Before this, developers made a manual setup.

How it works

A website that uses x401 tells the agent the proof it must give. The agent holds the task. The person gives approval with a biometric check, and a hardware key signs the correct terms. The business checks the signature offline, with no help. If one term changes, the signature is not correct.

The numbers

Proof gave no test results and no user numbers. Pat Kinsel, the head of Proof, tells that hundreds of millions of persons will first use ChatGPT to give an agent a task of value. That is a claim, not a measure.

What is not known

The source does not tell the number of businesses that use x401 at this time. Proof tells that stores and banks cause a blockage for agents. A password or a key does not show that a person approved one task.

Insight: Proof follows Adobe into the chat apps

Proof first had a manual setup. At this time it is in the ChatGPT app directory. On 4 November 2026, Adobe also put its marketing agent in ChatGPT, Claude and Copilot. This reads as a race to be in the large chat apps: the first approval standard that users hold can set the rules. Businesses that stop agents will possibly read x401 or a different standard.

Sources

  1. Proof launches its MCP in ChatGPT, starting with x401 so AI agents can prove who they act for and what that person approvedAiThority
AI MATTER · NEWS · AI MATTER · NEWS ·19 NOV2026

Posted

Tags

Learn the terms in this story

Guides for this story

More in Security

All Security news