Proof-of-Control v1.0 draft released for public comment

Free Instant SEO Audit

Glowing AI chip encased in a transparent cube demonstrating proof-of-control verification

By NewSunSEO September 19, 2026

Autonomous AI agents can now be held to a continuously verifiable standard. The Proof-of-Control v1.0 working draft, built with more than 80 global security leaders and hosted by the Linux Foundation’s decentralized trust arm, is open for public comment through October 30, 2026. Public launch briefings are scheduled for September 23, 2026.

What Proof-of-Control v1.0 proposes

Proof-of-Control is designed as an open standard for verifying what an AI agent is authorized to do and what it actually did, in milliseconds, without waiting for an annual or quarterly audit. It requires an agent’s authority to be defined before it acts, produces evidence at runtime that it stayed inside that authority, and makes that evidence auditable across the full agent lifecycle.

The v1.0 draft draws on production-proven verifiable AI technologies from member companies and was co-designed alongside a coalition of more than 80 security leaders across enterprise sectors. Public comment runs through October 30, 2026, with the working draft published at advancedaisociety.org/proof-of-control.

Why the standard is being housed at LF Decentralized Trust

The alliance behind the effort has joined the Linux Foundation and is hosting Proof-of-Control at LF Decentralized Trust as a lab. The structure is meant to keep verification infrastructure in a neutral commons, so that no single vendor owns the referee for what an agent did.

Foundational projects at LF Decentralized Trust, including AnonCreds and Trust Over IP, supply the cryptographic building blocks the standard adapts for the agentic stack.

How the standard is meant to work in practice

Three properties sit at the center of the v1.0 draft:

Authority must be defined before the agent acts, not after.
Runtime evidence must show the agent stayed inside that authority.
The evidence must be auditable across the entire agent lifecycle, by parties outside the operator.

The framing in the draft is that verification cannot belong to the verified. Unaffiliated, neutral third parties, alongside agent developers and deployers, must be able to openly verify the work of AI agents.

The wider context: a verifiability gap and pending rules

The release lands alongside bipartisan Congressional proposals that call for continuous agent verification and third-party evaluation, and as frontier lab leaders urge measured pacing of AI capability growth. The alliance describes a widening verifiability gap as autonomous AI agents begin to act faster than manual human oversight can follow.

Industry voices backing the effort

Supporters quoted around the launch include figures from enterprise security, cryptography, policy, and blockchain infrastructure:

Charles Iheagwara, Global Head of AI Cybersecurity at AstraZeneca, on the alliance’s Distinguished Review Board, framed the standard as a response to vendor assertions and called for transparent, inspectable ways to confirm what agents actually did.
Cryptographer Bruce Schneier, a Technical Advisor on the Distinguished Review Board, said that as agents become more powerful and ubiquitous, knowing who they belong to and what they are authorized to do will be critical.
J. Christopher Giancarlo, former Chairman of the U.S. Commodity Futures Trading Commission, described the standard as a way to give the agentic economy audit-ready evidence institutions can rely on.
Clay Shirky, Vice Provost for AI and Technology in Education at New York University, argued that human-in-the-loop is too generic and too vaguely specified, and that vendor self-regulation is too compromised.
Bob Blessing-Hartley, CTO of Shielded Technologies, tied the standard to the premise that users should not have to take anyone’s word for what an agent did.
Anoop Nannra, co-founder and CEO of Oko Labs, said the standard proves control at the network layer so machine risk becomes something the market can measure, price, and insure at agentic velocity.
Erynn Petersen, CEO of Emme, said the standard provides open, continuous ways to prove what an agent was authorized to do and what it actually did in healthcare contexts.
Patrick Duffy, Founder and CEO of Solv Labs, said institutions need defined authority and evidence they can check, not a persuasive explanation.
Noah Ringler, former AI Policy Lead at the U.S. Department of Homeland Security, called open verification a national security imperative when agents operate inside critical infrastructure.

Public launch event

A public launch briefing, “Who’s Watching the Machines? Launching Open Verification, the Proof-of-Control Standard for the Agentic Era,” is co-hosted by the alliance and LF Decentralized Trust on Wednesday, September 23, 2026. The program features alliance leaders, builders, deployers, and policy experts.

FAQ

What is Proof-of-Control v1.0?

It is a working draft of an open standard that requires an AI agent’s authority to be defined before it acts, produces runtime evidence that it stayed inside that authority, and makes that evidence auditable across the full agent lifecycle. The draft is open for public comment through October 30, 2026.

Who is hosting the Proof-of-Control standard?

The standard is hosted at LF Decentralized Trust as a lab, under the Linux Foundation, so that the verification infrastructure sits in a neutral commons and no single vendor owns the referee.

Why is open verification needed for AI agents?

Because AI agents can act in milliseconds, manual human inspection and periodic audits cannot keep pace. The alliance frames this as a verifiability gap, with bipartisan Congressional proposals now calling for continuous agent verification and third-party evaluation.