Skip to content

Proof-of-Control Standard

Can I deploy this agent and still account for what it does? How do I know the controls held?

Proof-of-Control is an industry-led open standard for evidence that an AI agent stayed within the controls it was given.

Hosted as a lab at LF Decentralized Trust

43days left in the public review of Proof-of-Control v1.0

Proof-of-Control v1.0 is open for live commentary through December 31, 2026.

Comment on the Google Doc →Join as Founding Contributor

The answer: openly verify agent controls.

Proof-of-Control is an industry-led open standard for evidence that an AI agent stayed within the controls it was given. Anyone can openly verify that evidence without trusting the operator or an auditor, and any change made after it is recorded can be detected.

Become a founding contributor → (Google sign-in required)



More than 100security leaders

have signed on to develop Proof-of-Control with our members.

See everyone who has joined ↓
127requirements defined

across six domains of verification.


Proof-of-Control solves for the Verifiability Gap

As machine capabilities compound exponentially and human oversight of machines scales linearly, the Verifiability Gap is the widening distance between what AI agents do and anyone’s ability to openly verify that they stayed within the controls they were given.

The Verifiability Gap in AI: when machine capability outpaces human oversight

CROSSOVER

VERIFIABILITY
GAP

HUMAN OVERSIGHT

MACHINE
CAPABILITY

CAPABILITY

TIME

Seeing what your agents achieved does not tell you what they actually did

Security leaders cannot bound an incident they cannot reconstruct.

Enterprises cannot show a board what their agents did last quarter.

Regulators cannot confirm that a high-risk system stayed within authorized parameters.

Insurers cannot underwrite what they cannot audit.

Procurement teams have no way to compare vendors on the one question that matters: can you show me what your system did, and can I verify it myself.

Nobody owns the action. A delegation chain runs three agents deep and no one can name who authorized the last step.

The enterprise reality of agentic AI

Every incident review begins with evidence the suspect produced.

You would never let a vendor audit itself and yet that is exactly what’s happening right now with your agents.


Enterprises need an alternative to launching agents on faith.

CISOs and business leaders are deploying agentic workflows blind. Enterprises are trapped between constraining agents into uselessness or unleashing them with unmanaged liability, leading to insurance exclusions, regulatory non-compliance, and unpriced liability.


AI assurance is insufficient for the enterprise agentic stack.

Agents are already reprogramming their own environments, acting on memory an attacker can poison, and minting identities for other agents.

  1. Before 01

    A policy states intent, not behavior.

    A governance policy describes what agents should do. It is not a record of what an agent actually did.

  2. During 02

    The system writes its own record.

    Logs come from the system you are investigating. If one step is compromised, its record can be too.

  3. Timing 03

    An audit sees what it samples, when it looks.

    Each audit is a person reviewing a sample of what already happened. What an agent does between audits, or outside the sample, goes unreviewed.

  4. After 04

    A contract assigns liability after the fact.

    To collect damages when something goes wrong, you need evidence of what the agent did. When the only record belongs to the vendor you are claiming against, how do you verify it?


Proof-of-Control supplies the assurance agentic systems are missing.

Proof-of-Control defines controls in advance and openly verifies every agentic action, without trusting the vendor.

  1. Before 01

    Policies can now be enforced by design and openly verified.

    Controls are defined before the agent acts, in a form a system can enforce. So far, we have mapped 127 requirements across Verifiability Tiers.

  2. During 02

    Evidence is produced as the agent acts.

    Alongside system logs, systems that conform to Proof-of-Control produce evidence of each action and whether it stayed within its controls, tied to a known operator. Any change made after it is recorded can be detected.

  3. Timing 03

    Every action is openly verified.

    Open verification runs on each action as it happens. At Proof-of-Control’s highest Verifiability Tier, verification is continuous: no action can run without producing evidence.

  4. After 04

    Accountability is clear at the moment of the action.

    When something goes wrong, Proof-of-Control has already produced the evidence for all parties to investigate. The vendor does not control the evidence and no one has to reconstruct what happened after the fact.

Every party reaches the same answer from the same record.


Proof-of-Control produces the same answer from the same record.

Today, every party demanding assurance invents their own process.

Your buyer sends a questionnaire, your insurer sends its own assessment, your regulator requests a filing, your certifier asks for access, and none of those records talk to each other. Under one open standard, the evidence your system produces at execution is the same evidence all of them can verify, without any of them trusting you or each other.

Proof-of-Control

  • Buyer
  • Insurer
  • Regulator
  • Certifier
  • Vendor

Proof-of-Control features

What the evidence is

What your agent did, not what it can do

Formal verification says what a system is bounded to do. Evaluations say how it behaved in a test. Proof-of-Control shows what your agent did in your deployment.

Binary, not scored

Maturity scores and risk ratings are subjective judgments. Proof-of-Control produces binary evidence: an action either occurred and stayed within assigned controls, or it did not. There are no gradients to negotiate and no weightings to dispute.

Deterministic, not probabilistic

AI models produce probabilistic outputs, but the verification record must be deterministic. Every verifier reaches the exact same conclusion from the same record, regardless of who asks or when. The model stays probabilistic; the verification of what it executed does not.

Machine verification, not human sampling

SOC 2 relies on human sampling, and no amount of sampling keeps up with agents that act a thousand times a week. Proof-of-Control evidence is verified by machine, so a million of your agent’s actions cost close to what one costs.

What you do not have to trust, or buy

Open mechanisms, not trusted third parties

An independent auditor is still a party you have to trust. Open verification anchors verification in mechanisms anyone can inspect, with every party it still relies on named, while your data, your prompts, and your models stay private. One question sorts any vendor: can I verify it without trusting you?

An open standard, not a proprietary product

Proof-of-Control specifies what the evidence must be, not the tool that produces it, so you assemble the mechanisms you already run, per domain, against the tier your risk calls for. A mechanism invented five years from now can still conform.

Tamper-evident records, not self-reported enforcement logs

Runtime gateways enforce policy, and the operator controls the logs. A gateway that was misconfigured, bypassed, or disabled still produces a clean log. Proof-of-Control requires evidence written where the agent cannot reach it, and tamper-evident, so anyone holding the record can detect a change to it.

One verifiable record, not fragmented reporting

Your auditor, your insurer, your regulator, and your counterparty each ask a version of the same question, and today each answer is produced separately, on its own schedule. Evidence anyone can openly verify answers all four at once, and none of them has to take it on your word.


Proof-of-Control’s Verifiability Tiers:The Anchoring Framework of Proof-of-Control

The Verifiability Tiers are a four-level framework in the Proof-of-Control standard for verifying an agent's control adherence, ranked by trust: from taking the operator's word to evidence anyone can verify without trusting the operator or an auditor. The tiers grade one thing: where the root of trust sits, and so how much you must trust a party to verify that an agent stayed within its controls.

Claims-based AI: where most AI is today
Proof-of-Control: Verifiability Tiers 3–4 · open verification

VERIFIABILITY TIER 1

Assertion

their word

VERIFIABILITY TIER 2

Attestation

a third party vouches

VERIFIABILITY TIER 3

Trust-minimized

anyone can openly verify

VERIFIABILITY TIER 4

Self-enforcing

no evidence, no action

Who you must trust
The operator
A third party or qualified auditor
The verification mechanism itself, open to anyone (mathematical or distributed assumptions)
The protocol itself: continuous constraints, anchored in external hardware roots of trust
How it is verified
Not verified; asserted
An auditor verifies, with privileged access
Anyone can openly verify, no privileged access
Execution is gated by cryptographic proofs; verification is continuous and automated
When it is verified
Never, or only when questioned
After the fact, sampled
After the fact, but by anyone
Before every execution interval: no proof, no write
On integrity break
Nothing: the system runs regardless
Detected in audit, after the fact
Detected by anyone, but the system can still run
MUST NOT execute until the integrity condition holds again
Accountability
Operator, self-attested
Named human + institution, bound to policy
Same authority, externally verifiable without trusting the operator or their auditors
Accountable person whose authorization is un-bypassable

THE BINARY THRESHOLD: Graded by who you must trust, not by whether cryptography is used. It indicates the phase change from human verification to mechanical verification, from a party’s word to a mechanism anyone can openly verify.

Verifiability Tiers in Practice.

Verifiability Tier 1 · Assertion

“Show me your system prompt and your policy settings.”

If the operator’s own records are your only evidence, you have an assertion, not control. Treat it as unverified.

Trust required: total.

Verifiability Tier 2 · Attestation

“Show me your SOC 2 or ISO 42001 report.”

Good hygiene, but static and retrospective, and it cannot stop a live action. Necessary, not sufficient.

Trust required: the auditor.

Verifiability Tiers 3 → 4 · Trust-minimized, into Self-enforcing

“Show me the signed, openly verifiable evidence for this specific action, and show that policy runs in a trusted execution environment that halts on violation.”

Now you verify the mechanism, not the vendor. This is the bar.

Trust required: the mechanism and the parties it rests on, all named in the disclosure.

Don’t trust, verify.


The six domains of verification

Provenance

Lineage & origin

Where an artifact, model, or piece of data came from, that it matches what it claims to be, and the custody chain linking origin to the action record.

Privacy

Confidentiality

What data was touched, and that the purpose matched the grant, evidenced without exposing the data being protected.

Portability

Cross-context control

That evidence and control move across vendors, platforms, and environments as claimed.

Authorization

Granted authority

That each action stayed within the authority granted, and what happened when it did not.

Identity

Attribution

Whose agent acted, who delegated the authority, and that the operator authorized it.

Security

Environment integrity

That the code ran unmodified, in the environment claimed, with access controls enforced.


The market forcing function

Built for insurance adoption

Priced risk turns verification into a business mandate.

Carriers cannot underwrite what they cannot audit. The Insurance Working Group, convened by Vidur Nayyar, brings carriers, reinsurers, and underwriters together to convert runtime verification into priceable risk models, claims-evidence frameworks, and a clear insurability classification.

Learn more →


Open verification

From Proof-of-Control to open verification

Bringing open source’s core principles into the agentic era.

Open source verified the code you ship. But as non-deterministic AI agents take autonomous action, code transparency is no longer enough. Open verification is a new category of open source: the open standards and specifications that let anyone verify what an AI system did, without trusting the operator or an auditor.

Claims-based

You are given a report

You have to accept it

It requires a relationship with the operator

It stops at your own boundary

When it is contested, you have their word

Open verification

You are given evidence

You can confirm it yourself

It requires no relationship at all

It crosses every boundary your agents do

When it is contested, you have the record

VERIFIABLE AI · A NEW CATEGORY OF AI

OPEN VERIFICATION · A NEW CATEGORY OF OPEN SOURCE

Proof-of-Control · our flagship standard

Open verification is a new category of open source: the open standards and specifications that let anyone verify what an AI system did, without trusting the operator or an auditor.

Verifiable AI is a new category of AI: AI whose actions can be verified from evidence, built on open verification.

Proof-of-Control is our flagship standard for evidence that an AI agent stayed within the controls it was given. It is one of many ways to achieve open verification.


FAQ about Proof-of-Control

Each answer ends with links to the matching sections on GitHub. Read through all questions on GitHub →


Founding Contributors

Join as Founding Contributor

Distinguished Review Board

Its appointed reviewers vote and give final sign-off on each version. Member reviewers take part without a vote.

Michelle Dennedy

Michelle Dennedy

LinkedIn ↗

Creator of “Chief Privacy Officer”

Charles Iheagwara

Charles Iheagwara

LinkedIn ↗

Global Head, AI & Cybersecurity, AstraZeneca

Vinh Nguyen

Vinh Nguyen

LinkedIn ↗

Founder, Aligned Intelligence Advisory

Antonino Rotolo

Antonino Rotolo

LinkedIn ↗

Director of ALMA AI, University of Bologna; Professor of legal theory and AI & Law; Coordinator of EU Regulatory Sandboxes for AI

Bruce Schneier

Bruce Schneier

Website ↗

Security technologist

Amy Steagall

Amy Steagall

LinkedIn ↗

Stanford University Chief Information Security Officer

Leadership Team

Manish Bhatt

Manish Bhatt

LinkedIn ↗

Security Researcher, Amazon LEO

Bob Blessing-Hartley

Bob Blessing-Hartley

LinkedIn ↗

CTO, Shielded Technology

Stan Byers

Stan Byers

LinkedIn ↗

Vice President of Business Development and Transformation, VIA

Rosalyn Curato

Rosalyn Curato

LinkedIn ↗

Chief Innovation Officer & GM, Agentic Security

Patrick Duffy

Patrick Duffy

LinkedIn ↗

CEO, Solv Labs

Mostafa Elkordy

Mostafa Elkordy

LinkedIn ↗

Senior Cybersecurity Officer, UNICC

Craig Ellrod

Craig Ellrod

LinkedIn ↗

Founder, CEO of the HACKERverse®

Cristin Flynn Goodwin

Cristin Flynn Goodwin

LinkedIn ↗

Managing Partner, Advanced Cyber Law

Bhavya Gupta

Bhavya Gupta

LinkedIn ↗

Information Security Officer, Lead - Vulnerability Management, Stanford University

Dylan Hobbs

Dylan Hobbs

LinkedIn ↗

Principal Founding Engineer, Vouched. KYA-OS Author, Decentralized Identity Foundation (DIF).

Dinesh Mendhe

Dinesh Mendhe

LinkedIn ↗

Founder & Chairman, The AI Lab

Anoop Nannra

Anoop Nannra

LinkedIn ↗

Co-founder and CEO, Oko Labs

Sam Nathans

Sam Nathans

LinkedIn ↗

Blockchain systems architect, Boston Blockchain Association

Advait Patel

Advait Patel

LinkedIn ↗

Senior Site Reliability Engineer, Creator of OWASP DockSec, Founding Member of AIVSS

Drummond Reed

Drummond Reed

LinkedIn ↗

Director, First Person Cooperative

Mo Sadek

Mo Sadek

LinkedIn ↗

Technical Director @ Alice

Ed Sewell

Ed Sewell

LinkedIn ↗

Founder and Chair

Founding Contributors

Toufic Adlouni

Toufic Adlouni

LinkedIn ↗

Managing Partner, Renno & Co

Alice Albl

Alice Albl

LinkedIn ↗

Emergent Technology Researcher, Mueller Consulting

Raj Annaladasu

Raj Annaladasu

LinkedIn ↗

Staff Customer Success Engineer, Palo Alto Networks

Prasanna Bableshwar

Prasanna Bableshwar

LinkedIn ↗

Principal Software Engineer / Technical Lead, Atlassian

Michael Bain

Michael Bain

LinkedIn ↗

CEO, Silversight.ai / Member, BT6.gg

Abdelhamid Bakhta

Abdelhamid Bakhta

LinkedIn ↗

Head of Applied AI & Verifiable Intelligence

Tim Bansemer

Tim Bansemer

LinkedIn ↗

CEO inblock.io assets GmbH

Daveed Benjamin

Daveed Benjamin

LinkedIn ↗

Founder, Canopi

Ben Bentzin

Ben Bentzin

LinkedIn ↗

Associate Professor of Instruction

Adrian Birzu

Adrian Birzu

LinkedIn ↗

Observability and Security Black Belt - Distinguished Engineer Oracle

JP Bourget

JP Bourget

LinkedIn ↗

CEO, Blue Cycle LLC

Joe Braidwood

Joe Braidwood

LinkedIn ↗

Co-founder & CEO, Glacis Technologies, Inc.

Carole Wai-Hai Brenner

Carole Wai-Hai Brenner

LinkedIn ↗

Consultant, AI & Data applied to Go-To-Market for B2B SaaS. Expert for the European Commission

Mihaly Bujaki

Mihaly Bujaki

LinkedIn ↗

Security Manager, Deutsche Telekom Security

David Campbell

David Campbell

LinkedIn ↗

Head of AI Security, Scale AI

David Cass

David Cass

LinkedIn ↗

CISO and Enterprise Technology Leader, Harvard (HES) Faculty

Vivian Chan

Vivian Chan

LinkedIn ↗

COO, 5arz

Divya Chandana

Divya Chandana

LinkedIn ↗

Sr software developer

Sharath Chandra

Sharath Chandra

LinkedIn ↗

Founder & Architect

Ying-Jung Chen

Ying-Jung Chen

LinkedIn ↗

Machine Learning Engineer, Google

Apurva Chokshi

Apurva Chokshi

LinkedIn ↗

CEO, AI Cyber Alliance

Vinod Choyi

Vinod Choyi

LinkedIn ↗

Distinguished Engineer - Security Architect, Verizon

Ben Christensen

Ben Christensen

LinkedIn ↗

Head of Partnerships, AI 2030

Eugene Coffie

Eugene Coffie

LinkedIn ↗

CEO & Founder of Predict AI

Crystal Coindreau

Crystal Coindreau

LinkedIn ↗

A.V.P., Sr. Security Architect - AI

Andre Darville

Andre Darville

LinkedIn ↗

Founder & CEO, 5arz

Andrew Davis

Andrew Davis

LinkedIn ↗

Founder, Living Code

Carlos De La Vega

Carlos De La Vega

LinkedIn ↗

Consultant, Implementation Engineering, Imprivata. Founding Engineer, Nubiz

Pieter de Wet

Pieter de Wet

LinkedIn ↗

Founder

Shane Deconinck

Shane Deconinck

LinkedIn ↗

Founder, Above Beyond

Ron F. Del Rosario

Ron F. Del Rosario

LinkedIn ↗

VP, Head of AI Security, SAP Supply Chain Management (SCM), Founding Co-Lead, OWASP Agentic Security Initiative (ASI)

Brijesh Deo

Brijesh Deo

LinkedIn ↗

Director of Software Engineering, Oracle

Leon Derczynski

Leon Derczynski

LinkedIn ↗

Principal Scientist, NVIDIA Corporation; Professor, IT University of Copenhagen

Fraser Edwards

Fraser Edwards

LinkedIn ↗

CEO, cheqd; Board member, Ayra

Eric Forst

Eric Forst

LinkedIn ↗

Co-founder & CEO, Blocksee; Member, American Society for AI (ASFAI)

Angela Garabet

Angela Garabet

LinkedIn ↗

AI Safety Researcher & Builder

Becky Gietzel

Becky Gietzel

LinkedIn ↗

Principal Security Architect

David Goecke

David Goecke

LinkedIn ↗

Founder, gotech.ai

Rakesh Gohel

Rakesh Gohel

LinkedIn ↗

Founder, JUTEQ Inc. (agent governance and evidence infrastructure)

Shiva Kumar Gosula

Shiva Kumar Gosula

LinkedIn ↗

Information Security Engineer, FinMkt

Matt Grasser

Matt Grasser

LinkedIn ↗

CTO and Co-Founder, Digital Transformation Solutions

Dazza Greenwood

Dazza Greenwood

LinkedIn ↗

Founder, CIVICS.com and Interlateral.com

Nas Hajia

Nas Hajia

LinkedIn ↗

Security Architecture, Lam Research

Eugenio Herrera

Eugenio Herrera

LinkedIn ↗

Investment Associate, Grupo MRP

Joel Hillier

Joel Hillier

LinkedIn ↗

CEO and Founder

Sai Honig

Sai Honig

LinkedIn ↗

Responsible AI & CyberSecurity Leader | International Speaker on Security, AI Ethics & Women in Tech | Senior Security Consultant, Novera Consulting

Ahmer Inam

Ahmer Inam

LinkedIn ↗

Founder and CEO, Cognisee PBC

Wes Jackson

Wes Jackson

LinkedIn ↗

AI Specialist Solution Architect, Red Hat

John Jiang

John Jiang

LinkedIn ↗

Cloud and Security Architect

Michael Kantor

Michael Kantor

LinkedIn ↗

President, Hashgraph Online (HOL); Co-founder & CEO, Moonscape Labs Inc.; Technical Steering Committee member, Hiero

Rajesh Kanungo

Rajesh Kanungo

LinkedIn ↗

CEO of TalaSeure, Inc. CSA member

Diana Kelley

Diana Kelley

LinkedIn ↗

CISO, Noma Security

Ali Khan

Ali Khan

LinkedIn ↗

CEO, SHAPE, Visiting Professor

Najeed Khan

Najeed Khan

LinkedIn ↗

Founder & CEO, AgentV

Jim King

Jim King

LinkedIn ↗

CEO and Founder, IndagoX

Mahesh Kukreja

Mahesh Kukreja

LinkedIn ↗

Security Engineer, Zipline International

Markus Lam

Markus Lam

LinkedIn ↗

Co-founder, Angainor BD @Axal, in-coming president,NYU blockchain labs

Nelia Lanets

Nelia Lanets

LinkedIn ↗

Information Security Analyst & Program Manager, Stanford University

Nate Leaf

Nate Leaf

LinkedIn ↗

Co-Founder, RailMetrix Inc.

Nate Lee

Nate Lee

LinkedIn ↗

CEO, Trustmind.com, Executive-In-Residence, Scale Venture Partners, Founder, Cloudsec.ai

Scott Mace

Scott Mace

LinkedIn ↗

Freelance journalist

Ryan Mackie

Ryan Mackie

LinkedIn ↗

AI, Data Protection, Cyber Security & IP Lawyer, DPO, EU AI Act Specialist, AI Legal Consultant, AI Governance & AI and Algorithmic Audit and Risk Specialist | In-House & Forward-Deployed Counsel

Chris Mancuso

Chris Mancuso

LinkedIn ↗

COO & Co-Founder, VisiQ Labs

Tim Marple

Tim Marple

LinkedIn ↗

Co-Director, AI Research

Chris Marzilli

Chris Marzilli

LinkedIn ↗

Vice President of Customer Success, Advisor360

Lain McNeill

Lain McNeill

LinkedIn ↗

Founder, Hybrid Intelligent Systems Design & Integration (HISDI)

JoEl Meeks-Matous

JoEl Meeks-Matous

LinkedIn ↗

Founder and CEO, Fairhub Nexxus; Chief Strategy Officer, Montage Business Financial

Atsushi Mizushima

Atsushi Mizushima

LinkedIn ↗

Partner, Nishimura and Asahi

Dilip Mohapatra

Dilip Mohapatra

LinkedIn ↗

CEO

Rezza Moieni

Rezza Moieni

LinkedIn ↗

CTO and CyberSecurity Adjunct Lecturer

Vineeth Sai Narajala

Vineeth Sai Narajala

LinkedIn ↗

Senior Technical Leader - AI Security Research, Cisco

Andrew Nebus

Andrew Nebus

LinkedIn ↗

Cybersecurity, AI auditor, ex-CIO

David Ngene

David Ngene

LinkedIn ↗

Co-founder, CEO & CTO, PacSpace

Pamela Norton

Pamela Norton

LinkedIn ↗

Sovereign Finance AI-Architect, M5 Capital

Paul Oakes

Paul Oakes

LinkedIn ↗

Senior Enterprise Security Architect, TD

Pushpendra Pal

Pushpendra Pal

LinkedIn ↗

CEO, DapplePot

Govindaraj Palanisamy

Govindaraj Palanisamy

LinkedIn ↗

Principal Architect, Global Payments

Tymofii Pidlisnyi

Tymofii Pidlisnyi

LinkedIn ↗

Protocol Author, Agent Passport System (APS)

Brigitte Piniewski

Brigitte Piniewski

LinkedIn ↗

Healthcare Ambassador

Satnam Purewal

Satnam Purewal

LinkedIn ↗

ISACA Puget Sound, Board Member, AI Safety WG (Cloud Security Alliance)

Arvind Raja

Arvind Raja

LinkedIn ↗

Founder & CEO, Kurral

Sharath Rajasekar

Sharath Rajasekar

LinkedIn ↗

CEO, Co-Founder

Fabio F. Ramos

Fabio F. Ramos

LinkedIn ↗

Senior Director, Product, Infoblox

Marianna Richardson

Marianna Richardson

LinkedIn ↗

Director of Communication, G20 Interfaith Forum Assoc.

Kristian Ronn

Kristian Ronn

LinkedIn ↗

CEO, Lucid Computing

Nelson Rosario

Nelson Rosario

LinkedIn ↗

Partner @ Rosario Tech Law | Board @ Illinois Blockchain Association | AI Steering Committee @ ISBA | Advisor @ Advanced AI Society

Lise Ross

Lise Ross

LinkedIn ↗

Fractional GRC Consulting, Checkbox Compliance

Fernando Rosso Camacho

Fernando Rosso Camacho

LinkedIn ↗

Co-founder & CEO, Biosphere Rocks

Andrew Rubinger

Andrew Rubinger

LinkedIn ↗

Founder, withaileron.ai

Sheera Sabur

Sheera Sabur

Website ↗

Founder, CEO Fascil Solutions

Aman Sardana

Aman Sardana

LinkedIn ↗

Senior Manager (IC) - Architecture and Engineering, Global Payment Network Technology, Capital One

Udbhav Saxena

Udbhav Saxena

LinkedIn ↗

Software Intern at Vouched

Tal Shapira

Tal Shapira

LinkedIn ↗

Co-founder & CTO, Reco

Phil Stafford

Phil Stafford

LinkedIn ↗

Co-Founder & CTO, Veraidyn, PBC

Mikayla Stewart

Mikayla Stewart

LinkedIn ↗

Cofounder, Coldstart AI, Founder @ Atono, CTO & Cofounder @ Athena Collective

Dan Stocker

Dan Stocker

LinkedIn ↗

Executive Director, AI Security, JPMorganChase

Nowa Sutaka

Nowa Sutaka

LinkedIn ↗

CEO & Founder, Puddin AI

Ravi Tanguturi

Ravi Tanguturi

LinkedIn ↗

Chief AI Security Architect, Director - Solutions Engineering, PointGuardAI

Mikko Tarkiainen

Mikko Tarkiainen

LinkedIn ↗

Independent Builder — AI Agent Control & Verification

Don Templeman

Don Templeman

LinkedIn ↗

Founder, Aemula

Ramesh Thiagarajan

Ramesh Thiagarajan

LinkedIn ↗

Security Architect, Amazon

Mitchell Travers

Mitchell Travers

LinkedIn ↗

Soulbis and Agentprivacy founder, BGIN - Co-chair IKP Working group, Trust over IP LDFT - ZKP Task force co-chair

Yogesh Trivedi

Yogesh Trivedi

LinkedIn ↗

Founder and CEO, CognitivTrust

Julie Tsai

Julie Tsai

LinkedIn ↗

Board Member, Bay Area CSO Council; CISO-in-Residence Ballistic Ventures; Founder Polaris Collective; Cyber Co-Lead AI Insiders; IANS Faculty

John V

John V

LinkedIn ↗

AI red team SME at the Institute of Security and Technology (NC3)

Samuel Vance-Law

Samuel Vance-Law

LinkedIn ↗

Head of Research, Decentralization Research Center

Joshua Waldrep

Joshua Waldrep

LinkedIn ↗

Founder, Pipelab (Pipelock open-source agent firewall) Maintainer, Agent Evidence Levels

Lindsay Walker

Lindsay Walker

LinkedIn ↗

Product Manager, Hedera AI Studio

Aydan Wang

Aydan Wang

LinkedIn ↗

RNA Biology Researcher, University of Manitoba; Co-Founder, Angainor; Undergraduate Researcher, University of Pennsylvania

David Webb

David Webb

LinkedIn ↗

Critical Infrastructure Cybersecurity Advisor & Strategist, Cybersecurity and Infrastructure Security Agency

Caroline Wong

Caroline Wong

LinkedIn ↗

Chief Strategy Officer, Axari; Author, The AI Cybersecurity Handbook (Wiley, 2026)

Seref Yarar

Seref Yarar

LinkedIn ↗

Co-founder, Index Network

Santiago Zanella-Beguelin

Santiago Zanella-Beguelin

LinkedIn ↗

Principal Researcher, Microsoft


How to Get Involved

  • For organizations

    Join the Verifiable AI Consortium

    For start-up builders of verifiable AI or enterprises looking to adopt verifiable AI

    Member dues fund the category creation work: convenings, analyst-grade research, and industry working groups.

  • For individuals

    Join as a Founding Contributor for Proof-of-Control

    For engineering and security executives and practitioners who make decisions on how AI is implemented or purchased, including CIOs, CTOs, CISOs, security practitioners, cryptographers, GRC professionals and AI security researchers.

    There is no cost to join.

    Google sign-in required