Loading...
AI governance course
An AI governance course for the people who design the controls
A practical AI governance course for compliance, risk, data protection and operations leads, focused on standards, regulators and where the human checkpoint sits.
In focus →
Live routingHover to pause
Incoming
Checking…
AI handles0
A person decides0
Where the line sits
Governance is where good intentions become controls
Once AI moves from pilots into daily work, someone has to own the rules.
That usually falls to compliance, risk, data protection or operations leads, often on top of the day job.
Our AI governance course gives those people a working toolkit rather than a reading list.
It covers the main frameworks, what UK regulators expect and how to design human oversight that actually works.
It is practical responsible AI training, with ethics and risk treated as design decisions rather than slogans.
Where AI helps the governance team
- Drafting policy text and DPIA sections
- Building and updating the AI use-case register
- First-pass review of vendor questionnaires
- Collating incident and performance reports
What governance people must own
- Risk classification of each AI use
- Approval of high-risk uses and exceptions
- Designing who reviews, who overrides and who stops
- Data protection judgements and sign-off
- Reporting to leadership and regulators
Frameworks
An AI governance course grounded in real standards
Our AI governance training works through the frameworks your auditors, clients and regulators are most likely to mention.
ISO/IEC 42001 is the international standard for an AI management system, built on the familiar plan-do-check-act cycle.
The NIST AI Risk Management Framework is a voluntary US framework, widely used to structure AI risk work around govern, map, measure and manage.
DSIT’s introduction to AI assurance explains how UK organisations can show their AI is trustworthy.
We show where these overlap, so you build one set of controls rather than three.
- ISO/IEC 42001: management system, policy and roles
- NIST AI RMF: govern, map, measure, manage
- DSIT AI assurance: evidence and assurance techniques
- EU AI Act: risk tiers and human oversight duties
85%
AI acts alone
130
65% of 200 tasks
Goes to a person
70
35% of 200 tasks
Illustrative batch with a typical confidence spread. Your real split comes from a pilot on your own work. Some decisions go to a person whatever the score.
Regulators
Responsible AI training that starts with UK GDPR
For most UK firms, the first AI regulator they meet is the ICO.
Its guidance on AI and data protection covers lawful basis, fairness, transparency and individual rights.
The ICO says safeguards around significant automated decisions cannot be token gestures.
It also says human reviewers need training, and the authority to escalate or override the system.
The ICO notes this guidance is under review after the Data (Use and Access) Act 2025, so we teach from the current version and flag what may change.
Where you operate in the EU, the AI governance course also covers the AI Act, whose high-risk rules now apply from 2 December 2027 for Annex III systems.
AI drafts
Check
Person decides
Learns
01 / 04
Human oversight
Designing the checkpoint so it is more than a rubber stamp
Human oversight fails when reviewers see too many cases, too little context or have no power to say no.
We teach a practical design method: decide which outputs need review, who reviews them and what evidence they see.
We cover confidence thresholds, sampling, escalation routes and a clear stop button.
We also cover what to log, so you can show a client or regulator how a decision was reached.
This part of the AI governance course covers the ethics and risk side of AI in practical terms, rather than as abstract principles.
Many teams pair this course with an AI policy template and an independent AI assurance review.
Human oversight fails when reviewers see too many cases, too little context or have no power to say no.
How it works
Four steps, one owner
Each system ships with a named person accountable for what it does.
01
01
Baseline
We review your current AI use, policies and data protection documents before the course.
02
02
Teach the frameworks
Sessions cover ISO/IEC 42001, NIST AI RMF, ICO guidance and DSIT assurance, applied to your cases.
03
03
Design oversight
Participants design review, override and escalation for two or three of your real use cases.
04
04
Leave with artefacts
You leave with a draft risk register, oversight design and next steps for policy and assurance.
Questions
Asked and answered
Further reading
From the source
International Organization for StandardizationISO/IEC 42001:2023, AI management systems ↗NISTAI Risk Management Framework ↗Information Commissioner’s OfficeGuidance on AI and data protection ↗Information Commissioner’s OfficeHow do we ensure individual rights in our AI systems? ↗Why a rubber-stamp review is not meaningful human oversight.Department for Science, Innovation and TechnologyIntroduction to AI assurance ↗The UK government’s primer on assuring AI systems.European Commission AI Act Service DeskAI Act Service Desk: frequently asked questions ↗Current enforcement and high-risk application dates.
Keep reading
Related pages
GovernanceAI assurance →AI assurance, with evidence a person can stand behindGovernanceAI policy template →An AI policy template your team will actually followGovernanceAI readiness assessment →An AI readiness assessment that names the human checkpointTrainingAI training for executives →AI training for executives, because accountability does not delegate
Next step
Tell us the work. We’ll show you where the line goes.
A 30-minute call. We map one workflow, what the AI could take, and where a person must stay in the loop.