What Your Board Expects You to Have in Place on AI: Five Items That Control the Conversation

Executive Summary

• Board members at mid-market companies ($50M-$5B revenue, 200-5,000 employees) are now asking “what is our AI strategy?” — not because of fiduciary liability theory, but because their other portfolio companies are asking the same question and they expect a coherent answer. Only 26% of boards discuss AI at every meeting, but among organizations reporting high AI ROI, that figure jumps to 63% (Protiviti Global Board Governance Survey, n=772 board members and C-suite executives, Q4 2025).
• The gap between governance structure and operational readiness is severe. Among Fortune 500 companies, 70% have AI risk committees and 67% report infrastructure progress, yet only 14% say they are fully ready for deployment (Sedgwick, n=300 senior leaders, December 2025). At mid-market scale — where 92% encounter AI rollout challenges (RSM Middle Market AI Survey, n=966, 2025) — that readiness figure is almost certainly lower.
• Five documented items transform a CEO from reactive to prepared: a written AI policy, a completed risk assessment, an allocated budget with metrics, assigned accountability, and a measurement framework. The CEO who walks into a board meeting with these five items checked controls the conversation. The one who walks in without them answers questions instead of asking them.
• Among S&P 100 companies, only 28% disclose both board-level AI oversight and a formal AI policy (Harvard Law School Forum on Corporate Governance, citing Glass Lewis analysis, March 2026). At mid-market scale, this number is vanishingly small — which means the CEO who arrives prepared stands out.

The Board Question That Changed

Twelve months ago, AI was an agenda item boards tolerated. Today it is one they demand. The NACD 2026 Governance Outlook Survey (n=340-366 directors, October-November 2025) finds 62% of director respondents now set aside agenda time for full-board AI discussions. PwC’s 2025 Annual Corporate Directors Survey reports AI is the number one oversight area directors say needs more boardroom attention.

But there is a critical distinction: boards at companies with $50M-$5B in revenue are not asking the same questions as Fortune 100 boards. They are not debating AI committee charters or director skill matrices. They are asking simpler, sharper questions:

• Are we using AI? Where?
• Do we have a policy?
• What are we spending?
• Who is in charge?
• How will we know if it is working?

The CEO who can answer all five controls the meeting. The one who cannot will spend the next quarter answering follow-up questions instead of executing strategy.

Why Boards Are Asking Now

Three forces are converging that make this a 2026 problem, not a 2027 one.

Their other boards are asking. Directors who serve on multiple boards — the norm at mid-market scale — are hearing AI governance questions at every table. The Protiviti survey finds 63% of organizations reporting high AI ROI include AI on every board agenda. When a director sees structured AI governance at one company and silence at another, the silent company gets the uncomfortable question.

Investors are setting expectations. A Glass Lewis analysis finds 65% of U.S. investors believe companies should disclose board oversight of AI governance, and 49% want AI oversight codified in committee charters (Harvard Law School Forum on Corporate Governance, March 2026). This pressure currently affects public companies, but private-equity and venture-backed mid-market firms face the same scrutiny from their own investors.

The governance-reality gap is widening. Fortune reports that while formal AI structures proliferate, “processes, controls, tooling, and skills embedded in day-to-day work” remain underdeveloped (Fortune, December 2025). Boards that asked “do you have a plan?” in 2025 are now asking “what happened with the plan?” The shift from awareness to accountability is happening faster than most management teams anticipated.

The Five Items: What to Have Ready Before Your Next Board Meeting

1. A Written AI Policy Exists

Not a 40-page governance framework. A document that answers three questions: what AI tools are approved for use, what data can employees put into them, and what happens when someone violates the rules. Among S&P 100 companies, only 45% maintain a formal AI policy (Glass Lewis analysis, 2025). At mid-market scale, the number is far lower. EqualAI’s AI Governance Playbook for Boards — cited by WilmerHale in its January 2026 governance alert — positions a written policy as the foundation for every other governance action.

What to bring to the board: A one-page summary of the policy, the date it was approved, and the date it was last updated. If the policy does not exist yet, bring a timeline for creating one. Boards respect candor about gaps more than vague assurances about progress.

2. A Risk Assessment Is Complete

The board does not need a risk register with 200 line items. It needs to know that someone has systematically identified where AI creates exposure. Deloitte’s AI Governance Roadmap identifies risk assessment as one of six core governance areas and recommends organizations “establish risk appetite and tolerance levels specific to AI initiatives.” EY’s 2025 analysis of Fortune 100 proxy statements found 22% now flag AI hallucinations, inaccuracies, and bias as material risks — a number that was effectively zero two years ago.

For a mid-market company, the risk assessment should answer: Where is AI touching customer data? Where are employees using unapproved tools? What happens if an AI-generated deliverable contains errors? What is the financial exposure? These are not hypothetical questions. Credo AI’s Navrina Singh told Fortune that “shadow AI and unsanctioned tools” are proliferating across organizations that lack visibility into where AI operates.

What to bring to the board: A summary of the top three AI-related risks, how they are being mitigated, and what risks remain accepted. The board’s job is to evaluate risk appetite, not eliminate risk.

3. A Budget Is Allocated and Tracked

NACD’s 2026 survey finds 76% of directors indicate AI investments will “definitely” or “probably” factor into 2026 growth strategy. Deloitte reports 78% of organizations plan to increase AI spending in the next fiscal year. The board expects AI to be a line item, not a discretionary experiment buried in IT’s operating budget.

The budget question is not “how much are you spending on AI?” It is “do you know how much you are spending on AI?” At most mid-market companies, AI costs are scattered across Microsoft 365 Copilot licenses ($30/user/month whether used or not), CRM add-ons, department-level tool subscriptions, and shadow spending on individual ChatGPT accounts. The CEO who can present a consolidated view of AI expenditure — even if the number is modest — demonstrates financial discipline that boards value.

What to bring to the board: Total AI spend (including embedded AI features in existing tools), spend per employee, utilization rate on paid licenses, and projected spend for the next 12 months.

4. Someone Is Accountable

The EY Center for Board Matters finds that from January through November 2025, only 12% of Fortune 100 companies disclosed that board members received education or training on AI (EY proxy statement analysis, 80 Fortune 100 companies, 2025). At board level, that is a disclosure problem. At management level, the deeper problem is that 66% of directors globally report their boards have “limited to no knowledge or experience” with AI (McKinsey, 2025). When boards lack expertise, they need to know that someone in management has it.

Deloitte’s framework recommends organizations “identify leadership ownership of AI strategy” and “assign oversight responsibility to board level or specific committees.” At Fortune 500 scale, that means a Chief AI Officer or dedicated AI governance team. At mid-market scale, it means a named person — the CIO, the CTO, an IT director, or a department head — with explicit accountability for AI outcomes, a defined time allocation, and a reporting line to the CEO.

What to bring to the board: The name of the person responsible for AI, what they are responsible for, how much of their time is allocated, and what decisions they can make without CEO approval.

5. A Measurement Framework Is in Place

The Protiviti survey reveals a stark divide: 95% of high-ROI organizations express confidence in integrating AI operationally, versus 33% of low-ROI organizations. The difference is not spending levels — it is whether the organization measures what AI is doing. NACD finds 47% of directors say selecting AI tools that deliver ROI is a current challenge, and 32% identify uncertainty around AI’s ROI as the number one roadblock to adoption.

The board does not need a dashboard with 15 metrics. It needs three to five measures that connect AI activity to business outcomes: adoption rate (what percentage of licensed users are active), efficiency impact (has AI reduced time-to-completion on specific workflows), quality impact (error rates, rework rates), and cost trajectory (is spend growing faster than value).

What to bring to the board: Three to five AI metrics tracked monthly, a baseline from before AI deployment, and a candid assessment of what the data shows — including where results have been disappointing.

Key Data Points

What This Means for Your Organization

The distance between “we are working on AI” and “here is what we have in place” is the distance between a CEO who answers questions and a CEO who sets the agenda. Every data point in this analysis points to the same conclusion: boards are past the education phase and entering the accountability phase. The question is no longer whether AI belongs on the board agenda. It is whether management can demonstrate structured progress.

The five-item checklist is deliberately simple. A written policy, a completed risk assessment, an allocated budget, a named accountable person, and a measurement framework. None of these require a Chief AI Officer, a dedicated AI governance team, or a six-figure consulting engagement. They require a CEO willing to formalize what is probably already happening informally — and present it with the same rigor applied to financial reporting, cybersecurity posture, or succession planning.

The advantage for mid-market CEOs is timing. At S&P 100 companies, 45% already have formal AI policies and 54% disclose board-level AI oversight. At the mid-market, these numbers are far lower. The CEO who arrives at the next board meeting with all five items documented is not meeting a minimum standard — that CEO is ahead of peers. If this checklist raised questions specific to your organization, I would welcome the conversation — brandon@brandonsneider.com.

The CEO who controls the AI conversation at the board table earns something more valuable than a clean governance score: the trust to move faster. Boards that feel informed give management more latitude. Boards that feel uninformed add oversight. The five items on this checklist are not bureaucratic requirements. They are the price of operational freedom.

Sources

1. Protiviti Global Board Governance Survey — “Only 26% of Directors Discuss AI at Every Board Meeting” (n=772 board members and C-suite executives, Q4 2025). Independent survey — high credibility. https://www.protiviti.com/us-en/press-release-ai-board-meeting-discussions-global-survey

2. NACD 2026 Governance Outlook Survey — Board priorities, AI agenda time, growth strategy data (n=340-366 directors, October-November 2025). Independent — high credibility. https://www.nacdonline.org/all-governance/governance-resources/governance-research/outlook-and-challenges/2026-governance-outlook/boards-shift-their-focus-to-execution/

3. Harvard Law School Forum on Corporate Governance — “US AI Oversight Through Three Lenses: Investor Expectations, the S&P 100 and Company-Specific Analysis” (Glass Lewis analysis, March 2026). Independent analysis — high credibility. https://corpgov.law.harvard.edu/2026/03/11/us-ai-oversight-through-three-lenses-investor-expectations-the-sp-100-and-company-specific-analysis/

4. Fortune — “AI Governance Becomes a Board Mandate as Operational Reality Lags” (citing Sedgwick, n=300 senior leaders, December 2025). Journalism with industry survey — moderate-high credibility. https://fortune.com/2025/12/18/ai-governance-becomes-board-mandate-operational-reality-lags/

5. PwC 2025 Annual Corporate Directors Survey — AI as #1 oversight area needing more time, 35% board AI integration. Independent survey — high credibility. https://www.pwc.com/us/en/services/governance-insights-center/library/annual-corporate-directors-survey.html

6. Deloitte AI Board Governance Roadmap — Six core governance areas, 78% increasing AI spend, 76% expecting transformation. Consulting framework — moderate-high credibility. https://www.deloitte.com/us/en/programs/center-for-board-effectiveness/articles/board-of-directors-governance-framework-artificial-intelligence.html

7. EY Center for Board Matters — Proxy statement analysis: 12% Fortune 100 disclosed board AI training, 44% list AI in director qualifications, 22% flag AI hallucination risk (80 Fortune 100 companies, January-November 2025). Independent analysis — high credibility. https://www.ey.com/en_us/board-matters/board-oversight-of-ai

8. WilmerHale — “Board Oversight and Artificial Intelligence: Key Governance Priorities for 2026” (January 2026). Law firm analysis — high credibility. https://www.wilmerhale.com/en/insights/client-alerts/20260122-board-oversight-and-artificial-intelligence-key-governance-priorities-for-2026

9. RSM Middle Market AI Survey — 92% encounter AI rollout challenges, 70% need outside help, 39% cite lack of in-house expertise (n=966, 2025). Industry survey focused on mid-market — high credibility for target audience. Referenced from prior research.

10. McKinsey — “The AI Reckoning: How Boards Can Evolve” — 66% of directors report limited to no AI knowledge, board governance archetypes (2025). Consulting research — moderate-high credibility. https://www.mckinsey.com/capabilities/mckinsey-technology/our-insights/the-ai-reckoning-how-boards-can-evolve

Brandon Sneider | brandon@brandonsneider.com March 2026