--- title: "AI & the Board" date: 2026-07-12 description: AI changes how every Board duty is discharged, from director to Company Secretary, and moves none of the accountability. author: Mario Thomas canonical: https://mariothomas.com/briefings/ai-and-the-board/ --- ## Start here Two short reads before you go deeper: what this subject actually is, and the order to take it in. ### The Line That Does Not Move AI asks something different of every person who governs: the director, the Chair, the Chief Executive, the CFO, and the Company Secretary. The constitutional texts have not moved. Cadbury named collective responsibility in 1992, the Companies Act 2006 fixed the duties every director owes, and the FRC's 2024 Code restates the Chair's and the Board's responsibilities. Not one of those sources has been rewritten for AI, and none of them needs to be. What has changed, completely, are the conditions under which those duties are discharged. Most of the Boards I meet are getting this wrong in the same way. They treat AI as an agenda item that belongs to somebody else: the CTO, the awareness course, the one director with a technology background. Deloitte's 2025 Global Board Survey found two-thirds of Boards still report limited to no knowledge or experience with AI, and the pattern beneath that number is credentials standing in for capability. A Board that runs its packs through ChatGPT and has heard an external speaker can feel itself modern while remaining unable to challenge the AI the business actually runs. That is comfort without capability, and in a governance context it is worse than acknowledged ignorance. The position these articles take is that two remakings are under way at once: of the Board's own work, through the tools now composing packs, minutes, and preparation, and of the work the Board governs, through operational deployments most directors cannot interrogate. Through both remakings one line holds. Agency, the doing of specific work, can be transferred to a machine. Accountability cannot be transferred at all. Every seat at the table discharges that principle differently, which is why this briefing works through the roles one at a time rather than treating the Board as a single undifferentiated thing. Read the briefing and you should be able to make one judgement about your own boardroom: whether its engagement with AI is genuine capability or accumulated credentials, and where, specifically, accountability is transferring silently with agency. A director who can name that boundary can hold it. ### Five Seats, One Principle Start with AI and the Director. It is the playbook the rest of the series builds on, and it defines what directorial AI literacy consists of: not technical fluency, but four capacities that let a director interrogate maturity claims, assess whether governance is operational or merely presentational, identify material AI risk, and exercise independent judgement. Whatever seat you hold, that piece is the foundation. Then follow [the articles](#core-reading) around the table. The Chair polices the boundary between agency transfer and accountability transfer; the Chief Executive chooses the bets that matter and answers to the market for them; the CFO stands behind the numbers the machine now helps produce; the Company Secretary operates, in every pack and minute, the boundary the Chair polices. Each piece stands alone, so going straight to your own seat is a perfectly good route in. Read in sequence, they show one principle refracted through five different sets of duties. The remaining pieces turn the lens around: not AI as something the Board governs, but AI as something the Board can use. The 2022 article, The Board in the Machine, is where this body of work began, with five questions for any Board encountering AI in its business. The decision analytics articles and Maximum Fidelity set out how lagging, leading, predictive, and reasoned indicators give a Board the highest-fidelity view available before judgement is exercised, and how to build that capability without surrendering the judgement itself. The Great Remaking diagnostic gives a Board the questions to put to management about where its AI position stands, built on the data, talent, and process loops rather than pilot counts and budget lines, and a 2024 guest lecture at the London School of Economics records the case, made to an executive class, that organisational change with AI is led from the Board. Under all of it sit the pieces the series was built on. The Accountability Gap is where the principle every seat turns on was first set out, that agency can be transferred to a machine and accountability cannot; the Six Board Concerns are where the governance agenda the Chair refuses to let collapse into risk alone was first drawn; and The AI Talent Bifurcation follows the split between capability and credentials out of the workforce and into the Boardroom, the judgement the first essay asks a director to make about their own table. [Remake](#remake-assets) holds the named mechanisms beneath the thinking, worth having open as you read. And when you have finished, [the questions](#faqs) compress the briefing into what a director can ask at the next meeting. ## Core reading The playbook first, then one principle traced seat by seat: the Chair, the Chief Executive, the CFO, and the Company Secretary. 1. [AI and the Director: A Practical Playbook for Governing What You Can't Fully See](https://mariothomas.com/blog/director-ai-governance-playbook/) (11 minute read, 29 March 2026): Directorial AI literacy is not technical fluency. It is four specific capacities that let directors interrogate maturity claims, assess real governance, and exercise independent judgement. Podcast edition: 15 minute listen. 2. [AI and the Chair: Governing the Board Through The Great Remaking](https://mariothomas.com/blog/ai-board-director-chair/) (14 minute read, 26 April 2026): Existing chair responsibilities now require different execution as AI remakes both the Board's own work and the work the Board governs. Podcast edition: 16 minute listen. 3. [AI and the CEO: Choosing the Bets That Matter](https://mariothomas.com/blog/ai-board-director-ceo/) (12 minute read, 21 June 2026): AI can build, deliver, and draft, yet the chief executive still chooses and still answers. Accountability for the bets does not move. Podcast edition: 12 minute listen. 4. [AI and the CFO: Standing Behind the Numbers the Machine Produces](https://mariothomas.com/blog/ai-board-director-cfo/) (12 minute read, 7 June 2026): AI can run the close, sharpen the forecast, and operate out of sight, yet the CFO still signs. Accountability for the numbers does not move. Podcast edition: 13 minute listen. 5. [AI and the Company Secretary: Operating the Boundary the Chair Polices](https://mariothomas.com/blog/ai-board-director-secretary/) (14 minute read, 24 May 2026): Board packs, agendas, and minutes now reach directors composed by systems the secretary cannot fully interrogate. The chair polices the boundary; the secretary operates it. Podcast edition: 16 minute listen. ## Further reading - [The Board in the machine](https://mariothomas.com/blog/the-board-in-the-machine/) (10 minute read, 17 October 2022): AI and machine learning are becoming ubiquitous in business decisions, and Boards need to know what is deployed and how it is governed. - [Transforming the Board: Using Decision Analytics for Strategic Advantage](https://mariothomas.com/blog/board-ai-decision-analytics/) (13 minute read, 23 March 2025): Decision analytics moves the Board from backward-looking metrics to predictive indicators that model possible futures, and changes how directors exercise judgement. - [Maximum Fidelity: How Four Indicator Types Strengthen Board Decisions](https://mariothomas.com/blog/maximum-fidelity-four-indicators/) (13 minute read, 12 April 2026): Four indicator types give boards progressively higher decision fidelity: lagging, leading, predictive, and reasoned. Together they represent the most accountable governance instrument available. Podcast edition: 15 minute listen. - [Implementing Decision Analytics: A Practical Guide for Boards](https://mariothomas.com/blog/implementing-decision-analytics/) (11 minute read, 3 April 2025): Decision analytics at Board level works when staged: a parallel input before a replacement, human judgement in the loop while the record builds. - [The Great Remaking: The Questions Boards Should Be Asking About Their AI Position](https://mariothomas.com/blog/the-great-remaking-board-diagnostic/) (10 minute read, 22 March 2026): Pilot counts and budget lines cannot tell a Board whether work is being remade. These questions, built on the data, talent and process loops, can. Podcast edition: 14 minute listen. - [Harnessing AI for organisational change led from the Board](https://mariothomas.com/blog/london-school-economics-ai-board/) (5 minute read, 7 June 2024): A guest lecture at the London School of Economics on harnessing AI for organisational change led from the Board. - [The Accountability Gap: When AI Delegation Meets Human Responsibility](https://mariothomas.com/blog/ai-agency-accountability/) (15 minute read, 16 November 2025): Organisations are transferring decision-making agency to AI while accountability stays with people, and approving deployments without the verification capability that accountability needs. - [AI is transforming governance: Six key Boardroom priorities](https://mariothomas.com/blog/board-ai-governance-priorities/) (10 minute read, 4 February 2025): AI takes Boards from overseeing hundreds of decisions a day to millions a second, each needing to be transparent, explainable and correct: six priorities follow. - [The AI Talent Bifurcation: Are You Building Skills or Collecting Credentials?](https://mariothomas.com/blog/ai-workforce-bifurcation/) (8 minute read, 18 January 2026): Workers with real AI capability command premiums of 28-56%; those collecting credentials without it face a 29% penalty. The same split now reaches the Boardroom. Podcast edition: 12 minute listen. ## Remake The mechanisms beneath the thinking: the model, diagnostic, methodology, and principle from the Remake Library that turn this briefing into apparatus a Board can use. - **Model: Six Board Concerns**. An interconnected lens of six concerns, Strategic Alignment, Ethical and Legal Responsibility, Financial and Operational Impact, Risk Management, Stakeholder Confidence and Safeguarding Innovation, that must be orchestrated together so AI discussion does not collapse into risk management alone. [Remake Library](https://mariothomas.com/remake/library/six-board-concerns/) - **Diagnostic: Well-Advised Assessment**. The measurement instrument of the Well-Advised principle: an assessment of the value an AI investment is actually realising across the five strategic priorities, returning a balanced-value reading rather than a single ROI number. [Remake Library](https://mariothomas.com/remake/library/well-advised-assessment/) - **Methodology: AI Business Case**. The integrated decision framework that crystallises across an ADAPT engagement rather than at a single stage: strategic alignment established at Align, cost and readiness evidenced at Diagnose, value shaped at Advise, and execution designed at Plan. [Remake Library](https://mariothomas.com/remake/library/#ai-business-case) - **Principle: Minimum Lovable Governance**. Governance embedded in how work happens: proportionate to risk, continuous rather than episodic, and used because it works. [Remake Library](https://mariothomas.com/remake/library/minimum-lovable-governance/) ## Questions The questions directors ask me most often on this subject, answered from the work in the briefing. ### Do directors need to become technologists to govern AI? No. Directorial AI literacy is not transformer architecture or prompt engineering; it is four specific capacities: interrogating maturity claims, assessing whether governance is operational or merely presentational, identifying material AI risk, and exercising independent judgement. A two-day awareness course produces a credential, not a capability, and comfort without capability is worse than acknowledged ignorance. [AI and the Director](/blog/director-ai-governance-playbook/) sets out what the four capacities require. ### We have appointed a director with deep AI expertise. Is that enough? It helps, and it is not enough. Every director carries an equal share of collective accountability, so every director must engage; a Board that lets AI matters concentrate around one fluent seat has allowed the rest of the table to silently exit an obligation Cadbury imposed in 1992. Heidrick & Struggles' 2025 commentary and Russell Reynolds' 2026 analysis of 398 public company boards reach the same conclusion. The Chair's job is to refuse the proxy, as I argue in [AI and the Chair](/blog/ai-board-director-chair/). ### Should directors use AI tools in their own board preparation? Yes, with one discipline held absolutely. Agency, the doing of the work, can be transferred to a tool; accountability cannot. A director who reads only the AI summary of a 200-page pack and asks only the questions the AI generated has outsourced the role they were appointed to perform. The same applies to the secretariat: an AI-generated board pack is not a board pack. [AI and the Chair](/blog/ai-board-director-chair/) and [AI and the Company Secretary](/blog/ai-board-director-secretary/) work through the failure modes. ### When AI produces the forecast, the close, or the disclosure, who answers for it? The same people who always did. The CFO still signs accounts a machine helped produce, because accountability follows the consequence, not the line of sight. The Company Secretary still stands behind the legal record and the disclosures, however machine-assisted the drafting. Nothing in the Companies Act or the FRC Code moves the signature. [AI and the CFO](/blog/ai-board-director-cfo/) and [AI and the Company Secretary](/blog/ai-board-director-secretary/) trace where that line is being tested. ### Can AI improve the decisions the Board itself makes? Yes, and it is an opportunity many Boards overlook while they attend to the risk. Decision analytics moves a Board from backward-looking metrics to modelled futures, and the four indicator types, lagging, leading, predictive, and reasoned, together give directors maximum fidelity: everything knowable made available before judgement is exercised. The indicators inform the decision; they do not make it. Start with [Maximum Fidelity](/blog/maximum-fidelity-four-indicators/) and the [implementation guide](/blog/implementing-decision-analytics/). ### How often should AI be on our Board agenda? Often enough that the conversation never collapses into risk alone. Protiviti and BoardProspects' 2026 Global Board Governance Survey found only 26% of Boards discuss AI at every meeting, while 63% of high-ROI organisations do, against 13% of low-ROI ones. Cadence is the easy part; breadth is the discipline. The Six Board Concerns model gives the agenda its shape: Strategic Alignment, Ethical and Legal Responsibility, Financial and Operational Impact, Risk Management, Stakeholder Confidence, and Safeguarding Innovation. A Board that reaches only the fourth has discharged a fraction of the obligation, and the Chair's job is to keep the other five on the table, as [AI and the Chair](/blog/ai-board-director-chair/) argues. ### How do we tell AI governance that works from governance that exists on paper? By how management responds to challenge. Governance that works has been applied to real deployment decisions, tested against edge cases, and revised when it failed, and the people running it can cite the cases. Governance on paper produces vague reassurance, circular appeals to the policy document, and no example of a decision it ever changed. That is governance theatre: oversight in appearance only. The test asks nothing technical of a director, only the discipline to ask for the evidence. [AI and the Director](/blog/director-ai-governance-playbook/) sets out the questions, and the Minimum Lovable Governance principle explains why the smallest structure in use beats the most comprehensive one on a shelf. ### Is the Chief Executive or the CFO accountable for our AI investment decisions? Both, for different things, with the Board holding each to account. The Chief Executive chooses which bets the organisation makes and answers to the Board and the market for whether they paid off and for the truth of what was claimed about them. The CFO stewards the capital and answers for whether those bets earned their keep. Conflate the two and accountability blurs: a small 2025 KPMG survey of US finance and technology leaders found 59% of CFOs and 61% of CIOs each claiming primary responsibility for AI investment decisions. [AI and the CEO](/blog/ai-board-director-ceo/) and [AI and the CFO](/blog/ai-board-director-cfo/) draw the line between the choice and the stewardship. ## References The constitutional texts and current research these articles draw on, gathered for further reading. - **Cambridge Judge Business School** (1 December 1992): [The Cadbury Archive](https://www.jbs.cam.ac.uk/faculty-research/publications/cadbury-archive/). The Cadbury Report (1992) and its archive: the origin of the collective-accountability principle several pieces in this briefing return to. - **UK Government** (8 November 2006): [Companies Act 2006](https://www.legislation.gov.uk/ukpga/2006/46). The statutory duties every director owes, including the duty to exercise independent judgement. AI amends none of them. - **Financial Reporting Council** (22 January 2024): [UK Corporate Governance Code 2024](https://www.frc.org.uk/library/standards-codes-policy/corporate-governance/uk-corporate-governance-code/). The Code that fixes the responsibilities of chair, chief executive, and Board whose execution AI now changes. - **Institute of Directors** (January 2026): [NEDs Reimagined](https://www.iod.com/app/uploads/2026/01/FINAL-IoD-Business-Paper-NEDs-reimagined-14.01-6ca5096ee6348f2301347e942a1ffe29.pdf). The Commission report naming AI literacy, information overload, and eroding independence as challenges every NED must address. - **Institute of Directors** (2025): [AI Governance in the Boardroom](https://web.archive.org/web/20251121075957/https://www.iod.com/app/uploads/2025/09/AI-Governance-in-the-Boardroom-1c7612e872fa3fce3f9d6cad78b0b4ba.pdf). Survey evidence on directors using AI personally while lacking the governance infrastructure their oversight role requires. - **Deloitte** (28 April 2025): [Governance of AI: A critical imperative for today’s boards](https://www.deloitte.com/global/en/issues/trust/progress-on-ai-in-the-boardroom-but-room-to-accelerate.html). The scale of the literacy gap: two-thirds of boards report limited to no knowledge or experience with AI. - **Protiviti & BoardProspects** (18 March 2026): [How Boards Drive AI ROI: 2026 Governance Survey](https://www.protiviti.com/us-en/survey/global-board-governance-survey). Agenda discipline correlates with AI value: 63% of high-ROI organisations discuss AI at every meeting, against 13% of low-ROI peers. - **PwC** (19 January 2026): [PwC’s 29th Global CEO Survey: Leading through uncertainty in the age of AI](https://www.pwc.com/gx/en/issues/c-suite-insights/ceo-survey.html). Why choosing beats spending: 56% of chief executives report neither revenue gains nor cost reductions from AI. - **McKinsey & Company** (4 December 2025): [The AI reckoning: How boards can evolve](https://www.mckinsey.com/capabilities/mckinsey-technology/our-insights/the-ai-reckoning-how-boards-can-evolve). Board governance analysis built on survey data showing 66% of directors report limited to no knowledge or experience with AI. - **EUR-Lex** (12 July 2024): [Regulation (EU) 2024/1689 (Artificial Intelligence Act)](https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32024R1689). The Act's AI-literacy duty (in force February 2025) and the high-risk obligations, deferred in July 2026 to December 2027 and August 2028, that shape the disclosure duties the secretary assembles. - **Russell Reynolds Associates** (16 January 2026): [The New Qualified Technology Executive: Redefining Board-Level Technology Expertise in 2026](https://www.russellreynolds.com/en/insights/articles/the-new-qualified-technology-executive). Russell Reynolds' 2026 analysis of 398 public company boards and the new qualified technology executive, the data behind the proxy-director question. - **Heidrick & Struggles** (13 November 2025): [AI focus: How boards are finding expertise to chart the unknown](https://www.heidrick.com/en/insights/frontier-tech/ai-focus_how-boards-find-expertise). Heidrick & Struggles' 2025 commentary on how boards are finding AI expertise, read alongside the Russell Reynolds data. - **EY** (2 June 2026): [2026 EY Global DNA of the CFO Survey](https://www.ey.com/en_gl/insights/finance/dna-of-the-cfo-survey). EY's 2026 Global DNA of the CFO Survey, the evidence behind the CFO piece on stewarding AI capital. - **Heidrick & Struggles** (12 February 2026): [2026 CEO & Board Confidence Monitor](https://www.heidrick.com/en/insights/board-of-directors/ceo-and-board-confidence-monitor-2026). Heidrick & Struggles' 2026 CEO and Board Confidence Monitor, the survey behind Maximum Fidelity's reading of how far Boards trust the indicators in front of them. ## The ideas beneath this briefing Ideas I've named and matured writing about AI & the Board: what each one means, and where it started. - **Directorial AI Literacy**: Not technical fluency but four capacities: interrogating maturity claims, assessing governance adequacy, identifying material AI risk and exercising independent judgement on AI decisions a director cannot fully see. [Read more](https://mariothomas.com/blog/director-ai-governance-playbook/) - **Decision Fluency**: A chief executive's hands-on familiarity with AI tools sufficient to judge what a bet is worth, distinguished from coding skill and treated as a duty rather than a nicety. [Read more](https://mariothomas.com/blog/ai-board-director-ceo/) - **Minute-Fidelity Failure**: When AI-drafted minutes capture what was said but not what was contested, weighed or dissented from, leaving an official legal record no one can fully defend. [Read more](https://mariothomas.com/blog/ai-board-director-secretary/) - **Summary Substitution Failure**: When directors rely on an AI-generated summary of a board pack as their primary reading, applying statutory judgement to material whose framing and omissions the AI, not a human, chose. [Read more](https://mariothomas.com/blog/ai-board-director-secretary/) - **Decision Analytics**: AI applied to Board decision-making that models what could happen, evaluates responses, and quantifies outcomes across scenarios, integrating internal metrics with external signals rather than merely reporting historical performance. [Read more](https://mariothomas.com/blog/implementing-decision-analytics/) All concepts: https://mariothomas.com/glossary/concepts/ ## More Board Briefings More complete resources on AI and emerging technology for the Boards that need the full picture. - [AI Governance](https://mariothomas.com/briefings/ai-governance/): Governance people route around fails to govern; the task is governing AI the Board cannot fully see without strangling adoption. - [AI Accountability](https://mariothomas.com/briefings/ai-accountability/): Agency can move to the machine; accountability cannot, and answering for what AI decides now takes capability that policy alone does not supply. - [AI Strategy](https://mariothomas.com/briefings/ai-strategy/): Approving good AI projects is not a strategy, and the Board's move is from accumulating pilots to a strategy it owns. - [AI Risk](https://mariothomas.com/briefings/ai-risk/): The AI risks that bite are seldom on the register, and the bill for sovereignty shocks, readiness gaps, verification costs, and model risk arrives later.