How do you become an AI governance analyst?

An AI governance analyst connects what rules and frameworks require with what an organisation's agents actually do, and proves it with evidence. Most come from compliance, risk, audit, legal or data protection. To join them, learn the main frameworks and how governed agents work, earn a checkable certificate, and show you can map a requirement to a control.

What does an AI governance analyst do?

Checks that an organisation's AI is used within its rules and can prove it. For agents, that means questions such as: Who approved this agent's scope? What depth is the agent certified to, and is the certificate current? Which decisions are held for a person, and is that enough under the rules that apply? Can every decision be reconstructed from the record? What happens when an agent fails certification?

The analyst maps requirements to controls and controls to evidence, reviews incidents, prepares material for auditors and supervisors, and flags gaps. The fuller description of what an AI governance analyst does and the short definition of an AI governance analyst go further.

Which frameworks and rules should you learn?

Start with the ones most employers will mention, and learn them in substance.

Then add the rules for your sector and market. In banking, model risk is supervised in most markets, for example through SR 11-7 in the US, PRA SS1/23 in the UK and OSFI Guideline E-23 in Canada. This is an explanation of the rules, not legal advice; confirm how they apply to you with counsel.

What do you need to know about how agents work?

Enough to test controls against reality, not just documents. In a governed system:

• Each action needs a depth, from read (1) to draft (2) to routine action within bounds (3) to hard-to-undo actions (4). Depth is earned through certification on the organisation's own cases, and the lower of the role's ceiling and evidence wins.
• Certificates lapse after 90 days; a run that clears no depth drops the agent to depth 1 the same day.
• Every action meets five bounds at once, and the verdict names which one decided.
• Every verdict is filed in an append-only record, so decisions can be reconstructed.
• Facts carry provenance: sealed, signed, declared or external.

These features generate the evidence a governance analyst relies on. An analyst who can read a decision record and a certification result is worth far more than one who can only read a policy.

Which certificate should you hold?

Pryme Intelligence Academy's tracks build on each other. Certified Pryme Operator teaches how review and incident handling work on the real product. Certified Pryme Builder adds how agents are designed and certified. Certified Pryme Architect, for people who lead regulated, multi-agent deployments, needs about 30 hours, a written design and a board review with two Pryme assessors, and is valid for 24 months. A governance analyst usually starts with Operator and aims for Architect over time. Certificates are issued through VerifyDoc.ai and can be checked by employers.

What should you show an employer?

A requirement-to-evidence map for one agent, with sensitive details removed: the rule, the control, where the evidence lives and what you would check. Add an incident or gap you found and how it was closed. The AI governance analyst CV template shows a layout.

Which jobs should you aim for first?

Read adverts for duties: AI risk, model risk, AI compliance, controls testing, policy, assurance. Titles vary widely. Banks, insurers, large public bodies and providers of AI services all hire for this work. Market detail is in how to become an AI governance analyst in Canada and in Germany.

Where the Academy fits

Every Academy lab runs in a free Build workspace, so you can see decision records, certification results and lapsed certificates for yourself before you advise on them. Module tests are set in your own industry and role, and certificates are listed on a public register by name, credential and expiry, which is itself a useful example of checkable evidence.