Private technical workshop · Berkeley, California
A workshop on the technical tooling needed to verify claims about frontier AI development — what should be proven, to whom, and how.
Attendance by invitationWhy now
“Today, the world lacks the technical and governance tools to deliberately pace frontier-wide progress.”
Pacing the Frontier — a statement signed by 1,386 employees of frontier AI companies, July 2026
There is increasing agreement that independent assurance is valuable for verifying that AI developers' safety commitments are upheld and that risks from AI development are managed effectively. However, the pool of auditors available to perform these tasks is currently very small relative to the scale of relevant activity to be overseen within AI developers — a challenge likely to grow harder as the number of highly capable AI systems to be overseen increases and as AI automates a larger fraction of research and other types of work. It is unclear how independent verification might adapt to settings with limited trust between auditors and the companies they audit, or between auditors and governments, particularly in an international context such as potential US–China coordination on AI safety. Technical tooling that enables scalable, trustworthy, and privacy-preserving verification could help to address these challenges.
This workshop aims to bring together auditors, external researchers building verification mechanisms, and staff from AI developers to identify and prioritize what tooling needs to be developed.
Focus
The workshop centers on concrete claims that technical mechanisms could help establish — for example, which models are deployed, whether specified safeguards remain in place, how AI compute is being used, and whether agreed constraints on development or deployment are being followed. Participants will examine the feasibility and limits of existing approaches across relevant use-cases including internal governance and security, domestic auditing, and international coordination. They will stress-test proposals on near-term priorities and work toward a shared view of what to build next.
A short, shared list of claims about AI safety and pacing for which stronger verification would be most valuable, in domestic and international contexts.
Research questions and pilot projects that AI developers, external researchers, and other stakeholders could advance in the next 6–18 months.
Program at a glance
A detailed agenda will be shared with participants ahead of the event. The day is structured primarily around small-group discussion to enable participants to discuss the most important questions and to shape R&D priorities.
A framework of candidate claims and how they relate to existing commitments and oversight — then small groups critique it, surface what's missing, and nominate the claims that justify serious technical investment.
A landscape view of current AI workload verification research, and working sessions mapping claims to mechanisms: what evidence would establish each claim, where the weakest links are, and what defeats existing approaches.
Deliberately opinionated proposals for portfolios of priority projects, red-teamed in breakouts, then compared against shared criteria to sort projects based on their priority and fit for AI developers, external researchers or others to pursue.
A plenary to collect preliminary rankings, name the real disagreements, record the questions that could resolve them, and identify next steps.
Who's in the room
Around thirty participants, spanning the communities whose combined judgment the problem requires:
Research, infrastructure, security, and policy staff from leading AI companies working on or adjacent to verification.
Organizations building mechanisms for verifying claims about AI compute and workloads.
Third parties developing auditing and risk evaluation practices that verification infrastructure could support.
Philanthropic and institutional supporters positioned to back the projects the workshop prioritizes.
Format