Global Directors Council Roundtable
Global Directors Council Roundtable Co-hosted by Liat Ben-Zur and BoardOutlook.

Insights from Our Co-Host

"A few years ago, AI in the boardroom was primarily an innovation discussion, focused on how emerging technologies could enhance products, improve customer experience and unlock growth. That framing has now materially shifted. AI is increasingly embedded across almost every board-level agenda item, from strategy, cost structure and talent to risk, brand, M&A and succession. Few substantive board discussions today are untouched, in some way, by AI.

Yet while most directors now recognise that AI matters, far fewer have translated that recognition into practical implications for how boards operate. The conversation is still too often framed narrowly as a technology issue, rather than a broader shift in how decisions are made. The more fundamental question is no longer whether AI belongs in the boardroom, but how it should be embedded, on what terms, and with what consequences for those who fail to engage meaningfully.

A clear gap is emerging between awareness and application. Directors are increasingly interested not only in how AI shapes business strategy, but in how it can enhance their own effectiveness, from improving preparation and sharpening challenge, to strengthening pattern recognition and testing assumptions. This area, in particular, is where the appetite for learning is strongest across the director community.

Boards that will navigate this transition most effectively are unlikely to be those with the most technical expertise, but those who build sufficient fluency to ask better questions, interpret management responses with greater confidence, and recognise where opportunity or risk may be underweighted. This does not require every director to become a technologist, but it does require enough familiarity to govern credibly in a rapidly evolving environment. Increasingly, the most effective boards will be those that integrate AI into how they think, prepare and decide, rather than treating it as a discrete topic on the agenda.

The central risk for boards is therefore not simply missing a technological shift, but allowing a capability divide to emerge, between directors actively adapting their practices and those relying on legacy instincts in a context where intuition alone is becoming insufficient."

— Liat Ben-Zur

Executive Summary

AI is no longer a peripheral technology topic for boards. It is reshaping how organisations compete, how directors prepare, how risk is identified and how decisions are tested. Yet the gap between recognising AI's significance and integrating it meaningfully into board practice remains substantial and in many cases, is widening.

BoardOutlook recently convened a Global Directors Council roundtable bringing together non-executive directors and chairs from listed, private, government and not-for-profit organisations across Australia, North America and Europe. Participants brought direct experience from boards spanning mining, financial services, technology, real estate, energy, logistics, hospitality and higher education, alongside specialist expertise in AI governance, cybersecurity, regulatory affairs and behavioural assessment. The discussion drew on benchmarking data from board evaluations and skills matrices conducted across 150 boards globally to establish context before opening to candid, experience-based conversation under Chatham House Rules.

The data points in a clear direction. Over the last two years, BoardOutlook's data showed a significant upskilling among directors who self-identified as moving from general to advanced on technology and data capability. At the same time, not a single board in that 150-strong sample listed AI as an explicit skill on its matrix. AI appeared, when it appeared at all, as a sub-area under technology, data or digital oversight. The implication is that boards are responding to the AI shift in their personal capability, but have yet to formalise it as a governance competence in its own right. The space between those two facts, recognised by individuals, unaddressed by boards, sits at the heart of this paper.

Where AI Belongs in the Boardroom

Effective use of AI in the boardroom resists a single definition, but participants converged quickly on a useful framing. There are, in practice, three distinct conversations boards can have about AI, and they are frequently conflated.

The first is AI as a driver of company performance: how the business uses AI to grow, to operate more efficiently, to compete and to defend itself. The second is AI as a governance domain: what the board is accountable for in terms of AI risks, ethics, regulatory compliance and oversight of management's use. The third, and the one that received the most attention and the least prior investment from participants, is AI as a tool for directors themselves, to sharpen preparation, surface patterns, pressure-test management's narrative and discharge their own responsibilities more effectively.

Most boards spend considerable time on the first conversation, some structured time on the second and almost no deliberate time on the third. The appetite in the room for that third conversation was unmistakable. Several participants observed that they regularly use AI in their advisory and executive work, but barely use it at all in their roles as directors, a gap they recognised as both significant and largely unaddressed by their boards.

The framing that resonated most strongly was that AI's value in the boardroom lies in enhancing director judgement, not replacing it. Used well, AI can surface inconsistencies, compare current performance against prior commitments and model scenarios at a fidelity boards have not previously been able to achieve. It does not however, relieve directors of the obligation to read, to challenge or to decide.

The Widening Gap Between Adopters and Non-Adopters

The most emphatic theme of the discussion was the rapidly emerging gap at the level of individuals, boards and entire organisations between those leveraging AI and those not. Participants described the productivity differential not as incremental but as orders of magnitude. Companies moving quickly on AI are re-engineering workflows, shipping product and capturing efficiencies at a pace those that are not simply cannot match. The same dynamic, scaled down, was observed at the level of the individual director.

Examples were shared of senior directors on substantial boards who comment on AI strategy in board discussions without having engaged personally with even mainstream AI tools themselves. In one case described, a long-serving director responded with genuine surprise on being shown that AI could draft a memorandum from a brief. The disconnect was not malicious or deliberate; it was simply the natural consequence of a busy director relying on the framing they had absorbed from others rather than direct experience.

This pattern raises governance questions that go beyond personal preference. Where one or two directors lack the fluency to engage credibly with the issue, the burden of challenge falls disproportionately on those who can and the board's collective oversight is, in practice, narrower than its composition suggests. The cybersecurity debate of recent years offers a useful parallel. Boards largely concluded they did not need a dedicated cyber expert in every seat, but that every director needed enough fluency to ask the right questions. The same standard is now applying to AI, and the boards that recognise this earliest will be best positioned.

The external dimension also matters. Activist investors and other observers of board composition are increasingly attentive to visible signals of director readiness. A board biography that does not reflect ongoing engagement with AI, through either formal study, structured reading or direct experience is an increasingly conspicuous gap, and one that sophisticated external parties are beginning to notice.

The Security Paradox

A consistent and sometimes uncomfortable tension surfaced between directors' professional comfort with AI and their willingness to use it on board material. Several participants described being heavy users of AI tools in their advisory or executive work, while being unwilling to bring those same tools anywhere near their board papers. The reasons were consistent: the sensitivity of board documentation, uncertainty about data residency and retention, the risk that even a well-secured personal account might infer the company being analysed from incidental context, and the absence of formally sanctioned tooling at the board level.

The result is a paradox. Those most equipped to leverage AI in governance are often the most cautious about doing so in their board roles. Several directors described being comfortable using AI to analyse public information about their organisations — competitive landscape, public filings, media narratives — but drawing a firm line at any document marked confidential. Others described workarounds in which they describe their company in deliberately generic terms to extract useful AI analysis without exposing the underlying material.

There was a strong consensus that some uses should remain off-limits, at least under current conditions. Board minutes were widely seen as a red line. AI-generated recordings of board meetings were viewed similarly, with concerns raised about discovery exposure, the integrity of the formal record and the chilling effect such recordings can have on candid discussion. Some participants noted a clear emerging governance view that minutes should not be drafted by AI.

The view in the room was that this paradox resolves only when boards themselves take an active position and sanction specific tools. Without clarifying what may be processed where, hosting models within trusted infrastructure, and stating explicitly what is out of bounds, individual directors will continue to make individual decisions in the absence of guidance, with predictably uneven results.

Start with the Business Problem, Not the Technology

A strong perspective emerged that the most productive way to bring AI into board conversations is, paradoxically, to take the term out of the framing. The fundamental questions a board asks about competitive position, retention, capital allocation, growth, and emerging risk do not change because of AI. What changes is the toolkit available to address them.

Directors who lead with the business problem and surface AI as one means of addressing it tend to bring management with them. Those who lead with the technology often find themselves talking past executives focused on the underlying issue. Several participants described this as the single biggest unlock in their own conversations with management teams. Asking why a procurement workflow looks the way it does, or why legal contract review still takes weeks, opens a conversation about reinvention that an AI-led framing often closes prematurely.

This reframing also has implications for how boards monitor AI's strategic impact. Early-warning analysis, scenario modelling and competitive landscape assessment are areas where AI can extend a board's reach considerably, particularly when geopolitical or market events have multivariate effects across a portfolio. One participant described using AI on the day of a major geopolitical event to think through implications across each of the businesses on which they served. The value was not that AI produced the answer; it was that AI produced the structure within which directors could reason more effectively.

The Homogenisation Risk

A more cautionary theme concerned the long-term effect of AI on the quality of board conversation itself. Several participants expressed concern that heavy reliance on AI-generated content tends toward a regression to the mean. Strategy that sounds plausible but generic, analysis that lacks edge, and a dulling of the genuinely diverse thinking that often produces the most valuable boardroom moments.

The observation was made repeatedly that some of the best board contributions come from directors asking the unexpected question. The one that comes from left field, that no model would have generated, but that opens a line of inquiry the board cannot afterwards ignore. The risk is not that AI replaces directors, it is that AI subtly narrows the range of what gets considered, by drawing every conversation toward the centre of mass of its training data. In a domain where the unexpected challenge is often the most valuable, that narrowing is consequential.

The same risk was observed in adjacent settings. Participants with university experience described a measurable convergence in student questions and assignments since AI tools became widely available, with critical thinking increasingly something that has to be deliberately exercised rather than assumed. The implication for boards is similar. The most effective use of AI in the boardroom is to broaden the range of considerations, pressure-test assumptions and identify blind spots, not to draft the document, write the strategy or generate the question. AI as a sparring partner sharpens judgement. AI as a ghostwriter dulls it.

Practical Use Cases: What Is Actually Working

Several specific applications surfaced in the discussion that participants described as immediately and meaningfully useful. They are worth recording, both because they demonstrate AI's current value and because they sit firmly within the boundaries of considered governance.

• Pressure-testing draft material: in one example shared, a board member used AI to test the language of a proposed code-of-conduct revision against the company's stated values, surfacing a striking inconsistency between the warmth of the values and the command-and-control tone of the document. In another, a draft market release was pressure-tested for likely investor response, with adjustments to specific words materially shifting the predicted reaction.
• Querying institutional memory mid-meeting: where boards retain papers in a sanctioned environment, AI can be used to surface specific information rapidly during discussion, whether that be a prior commitment, a previous representation or a number quoted at a previous meeting that would otherwise require hours of manual searching. This was widely viewed as one of the highest-value, lowest-risk applications.
• AGM and stakeholder preparation: multiple participants described AI as genuinely useful in preparing chairs for AGMs, generating likely shareholder questions and stress-testing prepared responses. In at least one case, AI-generated question sets were judged more comprehensive than those produced through traditional investor relations processes.
• Synthesising large volumes of stakeholder feedback: in governance contexts involving high-volume input such as regulatory consultation responses or community submissions, AI's ability to identify themes, surface tone and quantify sentiment was described as a step-change in what is practically possible in a short period of time.
• Crisis support: in at least one scenario shared, AI was used during an acute governance situation to rapidly synthesise constitutional, contractual and policy documentation, surfacing key considerations in minutes rather than days. The point was not that AI made the decision; it was that AI gave the board the structured ground from which to reason under time pressure.

Practices That Build AI-Capable Boards

The discussion also surfaced a practical set of governance practices that distinguish boards making meaningful progress from those treating AI as either a future problem or someone else's responsibility.

• A structured framework for board use of AI: the most disciplined boards are beginning to think about AI use across the full board cycle. Preparation before the meeting, support during the meeting and reflection and action afterwards. Each phase raises distinct questions about what is appropriate, what is sanctioned and what is genuinely additive. Articulating this explicitly removes ambiguity and creates the conditions for consistent practice.
• AI as an explicit board skill: the finding that none of the 150 boards leveraging BoardOutlook's skills matrix listed AI as an explicit skill is a clear signal of an opportunity. Naming AI capability directly on the skills matrix and assessing it candidly is a near-term step that boards can take to surface gaps and address them deliberately.
• Sanctioned tooling and clear red lines: the security paradox resolves only when boards make explicit decisions about which tools may be used on board material, where data is hosted, how retention is handled, and what is out of bounds. Hosting models within the organisation's existing trusted cloud infrastructure was viewed as a meaningful additional safeguard, both substantively and in terms of director confidence.
• Pressure-testing rather than drafting: the strongest current use cases are not those where AI generates content, but those where AI challenges content already produced. Whether thats identifying blind spots, surfacing the questions an analyst or activist might ask, testing the consistency of a document against stated values or modelling likely market response. AI used this way sharpens judgement; AI used to draft tends to dull it.
• Visible continuing education: as external scrutiny of board readiness intensifies, directors who can demonstrate ongoing investment in understanding AI through formal study, structured reading or direct experience are better positioned than those who cannot.

How BoardOutlook Supports Boards on AI Readiness

AI readiness in the boardroom needs to be measured and actively managed, not assumed. BoardOutlook's platform supports boards across the full set of processes that make this possible: board, committee and director evaluations; skills matrices that can incorporate AI as an explicit competence; CEO and chair evaluations; and Director 360s that surface how directors are actually showing up alongside the credentials they bring.

Where this becomes particularly powerful is in combination with AI itself. BoardOutlook's AI powered tool, OutlookIQ enables boards to interact with their performance, composition and evaluation data dynamically rather than through static, point-in-time reports. Drawing on a board's own underlying data, BoardOutlook's global benchmarking dataset and an institutionalised governance framework, the platform supports thought-partnered analysis across succession planning, capability gap identification, performance oversight, risk and scenario planning, and strategic priority alignment. All of this operates within hosted, audited infrastructure with deletion and retention controls explicitly engineered for the sensitivity of board data.

For boards ready to take AI readiness seriously as a governance priority, both as a domain to oversee and as a capability to build into their own work, BoardOutlook provides the tools, data and structured process to make that commitment concrete and sustained.

For information on upcoming sessions or the BoardOutlook platform, please contact the BoardOutlook team at jingqi.wu@boardoutlook.com.

If you're interested in connecting with Liat Ben-Zur, you can learn more and get in touch via her website: lbzadvisory.com.

This paper was developed by BoardOutlook based on a Global Directors Council roundtable co-hosted with Liat Ben-Zur. Participant contributions are reflected thematically and are not individually attributed, in accordance with Chatham House Rule. BoardOutlook's Global Directors Council brings together chairs and directors from across sectors and geographies to explore governance questions in a confidential peer setting.