Who's Watching the Machines? Launching Proof-of-Control, an open verification standard for the agentic era
Advanced AI Society and Linux Foundation Decentralized Trust (LFDT) invite you to join us for our three-hour online event, "Who's Watching the Machines? Launching Proof-of-Control, an open verification standard for the agentic era." We are convening the founders and operators building the open verification ecosystem: open-source maintainers, enterprise security executives, insurers, and policymakers. September 23, 2026, at 8:00 PT, 11:00 ET, 17:00 CEST, and 23:00 SGT. Register for the zoom link via LFTD directly.
Why now
The AI debate reaching Congress this month is about what companies should disclose about their models before release. Every proposal on the table rests on what a company reports about itself. Open verification addresses what happens after: once an agent is deployed and acting, what evidence exists that anyone can verify without taking the operator's word. Enterprises, insurers and regulators are already carrying that risk, and no policy on the table reaches it.
At a glance
Open source won the enterprise on a security premise: you cannot secure what you cannot inspect. But while open-source code defines how a traditional system is built, non-deterministic AI agents make dynamic decisions at runtime.
Open verification is the next category of open source, extending the ethos from inspecting source code to allowing anyone to openly verify actions in the agentic stack. As the open standards framework for verifiable AI, open verification is itself open source, built on open standards anyone can implement, and moves the root of trust from a party to a mechanism anyone can verify.
Because the standards are open, verifiable AI protocols interoperate, so deployers are not locked to one vendor.
Openly verifiable evidence unlocks the defensible, enterprise-wide scaling of autonomous agents: creating competitive markets, reopening the path to AI insurability, and preserving public accountability as AI moves into critical real-world systems.
Advanced AI Society will introduce Proof-of-Control, the anchoring open verification standard and the newest community lab at LFDT. V1.0 is now open for comments.
Who should attend
Anyone trying to understand what verification means as AI agents begin making decisions about people's money, health care, and access to public services.
Enterprise AI leaders, CIOs and CTOs moving from static models to autonomous, non-deterministic agent workflows.
CISOs and security executives managing runtime risk, agentic privileges, and unauthorized execution.
Insurers, regulators and auditors setting compliance standards and defining liability for dynamic AI outputs.
Open-source contributors and cryptographers defining the standards stack for agentic identity, execution, and verification.
What you will learn
The category shift, from claims-based AI to verifiable AI. Why non-deterministic AI outruns claims-based systems, and why verifiable AI is a distinct technology category rather than a feature.
The open source evolution. How open verification extends the open-source argument from code inspection to runtime action verification. And why open-source code or model weights or open training data tell you nothing about what an agent actually did in production.
The anchoring standard. How Proof-of-Control and its core framework the Verifiability Tiers give security teams the foundations for designing the systems to openly verify agent behavior without trusting the operator.
The interoperable ecosystem. How standards bodies and market leaders are building cross-industry routes to verifiable agentic trust.
Part One — The category: why open verification, and why now
Learn how the open-source movement is expanding to address non-deterministic execution through open verification. Verifiable AI is emerging as a new category, made interoperable through open standards that focus squarely on agent behavior, runtime infrastructure, and production deployment. We then gather perspectives from AI Safety, Verifiable AI investors, insurance leaders, and the assurance industry to examine why unverified agent behavior has become a critical liability, and why this moment demands an open standard.
Talk: Welcome Address, Daniela Barbosa, Executive Director, LFDT
Talk: Who's watching the agents? What open verification is, the three qualifying properties, and why this moment. Tricia Wang, CEO, Advanced AI Society
Talk: Open verification as the next category of open source. Jim Zemlin, Executive Director, Linux Foundation
Panel: Agent behavior is a security problem. Moderated by Angel Saad Gomez, Advanced AI Society and Point 72
Panel: Why now, from three angles. Safety, insurance, assurance. Moderated by Bettina Warburg, Co-founder, Advanced AI Society
Part Two — Proof-of-Control: moving the root of trust
A deep dive into Proof-of-Control, the anchoring open standard of the new lab at LF Decentralized Trust, designed to move the root of trust off the system operator. We introduce its anchoring framework, the Verifiability Tiers, which force absolute clarity on where the root of trust is held across an agentic system. Through real-world use cases, high-profile agent failure scenarios, and direct conversations with enterprise leaders, we examine why major deployers are refusing to put agents into production without verifiable controls.
Talk: Introducing Proof-of-Control. Tricia Wang, CEO, Advanced AI Society
Demos: Live implementations of Proof-of-Control
Panel: The demand from deployers: We will not deploy what we cannot verify
Part Three — The ecosystem: many routes to open verification
Open verification is an expansive ecosystem rather than a walled garden. Discover how verifiable AI pioneers are implementing open standards across cryptographic primitives, zero-knowledge proofs, and sovereign compute infrastructure. We examine the emerging primitives and concrete sector deployments in healthcare and finance, before closing with a look at the emerging agentic stack being defined at the Agentic AI Foundation (AAIF) and the priorities of open verification research.
Hosted by Sheila Warren, Co-founder, Advanced AI Society, host of Part Three
Showcase: The emerging primitives
Showcase: The emerging sovereign stack
Showcase: The emerging health stack
Showcase: Finance
Talk: The agentic stack at the Agentic AI Foundation, Mazin Gilbert, CEO of AAIF
Talk: The research stack for open verification. Jim Schwoebel, Director of Research, Advanced AI Society
Builders and deployers hands-off with Proof-of-Control
11:00am to 11:30am PST 14:00 PST / 20:00 CEST For builders, researchers and maintainers looking to shape the standard: the state of the field and what research is still needed, a walkthrough of the GitHub repository and the six domain working groups, how to join the community on Slack, and the open questions that need the field's help. Public comment on the standard is open until 30 October, so there is time for what you bring.