AC2 Protocol: The missing security layer for AI agents
The AC2 Protocol is a new security layer designed to provide cryptographic proof of intent for AI agents, addressing the lack of human-in-the-loop verification in current frameworks. It allows users to sign actions via hardware-bound keys without sharing credentials with the agent.
Why it matters
As AI agents gain the ability to perform sensitive tasks like payments and code signing, verifiable identity and intent become critical for security.
Documentation AC2 Protocol Why AC2 Try AC2 Integrate AC2 Get Started Documentation AC2 Protocol Agentic Communication + Control Protocol (AC2) upgrades your stack with self-custodial signatures that ensure every action is a verifiable expression of your sovereign will, all without ever sharing your keys.
It's cryptographic control for the agentic age.
The gap in today's HITL frameworks As AI agents take on payments, code signing, and other sensitive operations, today's human-in-the-loop approaches lack verifiable identity, standardized delegation, and cryptographic proof of user intent. AC2 fills those gaps with hardware-bound approvals, credential isolation, and signed, defensible audit trails.
Native Chat HITL is the standard. AC2 is the sovereign upgrade.
Stop trusting chat button clicks and start owning your actions with cryptographic proof.
Instead of giving agents credentials, AC2 requests a hardware-bound signature from the user, generating cryptographic proof of intent before execution.
Already running an agent framework like OpenClaw?
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in