Confidential benchmarking
The benchmark compares how members staff, govern and test their AI oversight programs. Each member sees their own position against peers. It never shows a figure for a group of fewer than five.
AI accountability.
A shared discipline.
A confidential peer network for the people accountable for AI.
Request an invitationSummits
Why the Network exists
Every regulated financial institution now has a person who is accountable for AI.
Boards and supervisors ask this person how many staff the AI program needs, which controls will pass an examination, and how often AI systems fail. There is no reliable answer, because each institution keeps its data private and no one combines it.
The AI Risk & Compliance Network brings these people together. Its rules protect what members say and share, so they can speak openly and use real numbers.
What membership provides
Members share confidential data about their AI programs. In return, they get three things that vendors and analysts cannot supply.
Members-only roundtables
Members meet under the Chatham House Rule to discuss incidents, examinations, board reporting and the methods that worked. Vendors and press do not attend. We do not record the meetings.
A reference definition of the role
“The AI Governance Officer: A Reference Definition” describes the role and maps it to the NIST AI RMF, ISO/IEC 42001, the EU AI Act and Asian supervisory guidance. Members maintain it.
How contributions work
- 01
You contribute
You send confidential data about your AI program: staffing, governance, testing and incidents.
- 02
We combine it
We combine your answers with the answers of other members. We never show a figure for a group of fewer than five.
- 03
You see your position
You see privately how you compare with your peers. Other members see only the combined data, never your answers.
We never sell member data, share it or show it to sponsors.
No sponsor sees the member roster, member data or the benchmark. The same rule applies to every company that a convener is connected to. These rules are in the Network’s charter, which every member receives before they join.
Supervised
Supervised is the Network’s podcast about AI risk and the people who are accountable for it. The conveners host it. Guests will include regulators, chief compliance officers and general counsel. The first episodes come out in December 2026.
About SupervisedRequest an invitation
Founding membership is by invitation. The first cohort is small.
- 01Send a request with your role and institution.
- 02We send the charter and the reference definition.
- 03You join, contribute data and see your position.