Who Owns Board AI Education?

Most financial services boards now accept that they need to understand AI.

Fewer can say who is actually accountable for making that happen.

AI is typically raised informally in board discussions, often by whoever happens to be most technically confident — frequently the Chief Technology Officer or Chief Information Officer. That default is understandable, but it quietly reframes board AI education as a technology initiative rather than what it actually is: a governance and strategic capability issue.

Without a named owner, board AI education tends to happen unevenly, if at all, and becomes difficult to evidence when regulators or shareholders ask what the board has done to build its own AI literacy.

This article sets out who should realistically own board AI education, how that decision has traditionally been made for other emerging risk topics, and where AI-enabled tools can support — but never replace — that ownership.

Why ownership has become an urgent question

Boards are under growing pressure to demonstrate AI literacy, not just AI oversight.

Regulators, investors and staff increasingly expect boards to show they understand the risks and opportunities AI presents to the business, not simply that they have approved a policy document.

The UK Government's AI Skills agenda and wider regulatory direction of travel make this harder to defer. Boards that cannot evidence a structured approach to their own AI capability are exposed both reputationally and, increasingly, from a governance perspective.

The difficulty is that AI does not map neatly onto any single existing board responsibility. It touches risk, strategy, technology, operations and conduct simultaneously. That ambiguity is precisely why ownership so often drifts to whoever is loudest on the subject, rather than being deliberately assigned.

How boards have handled this before

This is not the first time boards have faced an emerging topic that did not fit existing structures.

Cyber security followed a similar path. Initially treated as an IT issue, most boards eventually recognised that cyber risk required board-level ownership — commonly landing with the Risk Committee Chair or a designated Non-Executive Director, supported by the CISO for technical delivery.

Climate followed a comparable pattern. Ownership settled at board or committee level, with sustainability or ESG leads handling delivery and reporting.

In both cases, the pattern was consistent:

• A board-level sponsor held accountability.
• An executive lead handled delivery and day-to-day management.
• Progress was reported through existing governance cycles rather than as a standalone initiative.

AI education is now following the same trajectory. The realistic ownership models are the Chair, the Senior Independent Director, or the Risk Committee Chair — supported by a Head of Learning and Development, Chief Risk Officer, or equivalent for delivery.

Where AI-enabled tools support the programme

Once ownership is assigned, AI-enabled tools can meaningfully reduce the effort involved in delivering and assessing board AI education.

Tailored learning content can be generated to reflect a board's specific sector exposure, rather than relying on generic training decks. Literacy assessments can be adapted to individual directors' starting knowledge, surfacing gaps more efficiently than a one-size-fits-all questionnaire. Progress tracking across a board or committee can be automated rather than manually compiled ahead of each review.

These tools are genuinely useful for scaling delivery and keeping content current as AI capability and regulation evolve.

What they cannot do is decide who is accountable, interpret what "good" board AI literacy looks like for a specific firm, or take responsibility when a gap is identified. That judgement remains squarely a board governance matter.

Assigning and documenting ownership in practice

Ownership should be treated as a governance decision with the same rigour as any other board accountability.

Practical steps include:

• Naming a single accountable board-level sponsor — not a committee collectively, and not "the board" in general terms.
• Recording this ownership explicitly in committee terms of reference or the board skills matrix.
• Separating accountability from delivery, so the sponsor is not expected to personally design or run training sessions.
• Building review of board AI literacy into existing governance cycles, such as the annual board effectiveness review, rather than treating it as a one-off project.

Leaving ownership unassigned carries real operational risk. Without a named sponsor, board AI education tends to be reactive, inconsistent between directors, and difficult to evidence if challenged by a regulator, auditor or shareholder. In an environment where AI literacy is increasingly expected rather than optional, that gap is no longer a minor administrative oversight.

Example

A London-based specialty insurer is preparing for its annual board effectiveness review. AI has come up informally at three board meetings that year, but no one is formally accountable for board AI literacy.

The Senior Independent Director raises this gap during the review.

The board responds by formally assigning ownership of board AI education to the Senior Independent Director, who sponsors a programme delivered by the Head of Learning and Development, with input from the Chief Risk Officer on regulatory context.

Progress is reviewed at each Risk Committee meeting and reported in the annual board effectiveness statement, giving the board a clear, evidenced record of its own AI capability development.

FAQs

• Should the Chief Technology Officer own board AI education?

  The CTO or CIO can be a valuable contributor, particularly on technical context and emerging capability, but ownership should sit with a board governance figure such as the Chair, Senior Independent Director or Risk Committee Chair. Assigning ownership to a technology leader risks framing board AI literacy as a purely technical issue rather than a governance and strategic capability responsibility.

• Can board AI education ownership be shared across multiple people?

  Delivery can and often should involve several contributors, such as L&D, the CRO and external providers. However, accountability works best when it rests with a single named sponsor. Diffuse ownership without one accountable individual tends to lead to inaction, since no one feels ultimately responsible for progress.

• How often should board AI education ownership be reviewed?

  Ownership and progress should be reviewed at least annually, ideally aligned with the existing board effectiveness review cycle. This keeps board AI literacy embedded in standard governance practice rather than treated as a separate, easily deprioritised initiative.