← Knowledge Base 🕐 24 min read
Knowledge Base

Board AI Strategy

· March 2026

The structured process by which a company's board of directors engages with AI strategy, risk, and investment oversig...

The structured process by which a company’s board of directors engages with AI strategy, risk, and investment oversight. As of April 2026, 79% of board members report limited-to-no AI knowledge (Deloitte, n=468, May-July 2024), yet 88% of their organizations deploy AI in at least one function. Closing this gap is now a fiduciary obligation, not a discretionary agenda item.

Why this matters to mid-market buyers

  • State AI laws and SEC enforcement require documented board engagement. Colorado, Texas, and California have passed AI-specific legislation. The SEC created CETU (Feb 2025) with AI-washing as a top priority. Boards that cannot demonstrate AI oversight face regulatory and investor pressure.
  • The 98/7 gap. A survey of mid-market CEOs via the Chief Executive Network found 98% see AI as critical, but only 7% have a defined strategy. Board-level engagement is the forcing function that converts awareness into structured action.
  • Proxy advisors are moving toward withhold recommendations for directors who cannot demonstrate AI literacy. 65% of U.S. investors now expect companies to disclose board oversight of AI governance (Sustainalytics, 2025).

Practitioner voices (pillar 13)

“98% of the CEOs said this is critical, I know that AI is going to impact my business. The flip side though, and probably what’s most interesting… 7% said, ‘Yeah, we have a strategy.’ So we all know it’s important, but we don’t really know what to do about it.”

— Chris Happ, CEO, Virtuous AI (Chief Executive Network survey) · April 2026 · research/13-multimodal-sources/ai-for-the-c-suite/2026-04-14-chris-happ-why-98-of-ceos-know-ai-matters-but-only-7-have-a-.md

“CEOs are looking at CIOs to be driving this strategy. What’s the vision? What is our strategy around AI? And how do we sort of go on this path? And there’s no other function than the CIO function to kind of drive this change.”

— Saket Srivastava, Chief Information Officer, Asana · September 2024 · research/13-multimodal-sources/me-myself-and-ai/2024-09-17-meet-your-new-teammate-ai-asanas-saket-srivastava.md

“A few years ago we had an AI workshop with board members. It was a full day AI workshop. […] One of the board members, she was a senior lady. She came to us and she was like, ‘I’m a little bit skeptical about AI.’ […] After the workshop, it was like an 8-hour session, she came to us and this is the exact sentence that she said, she was like, ‘You converted me.’”

— Dr. Sam Zolfagharian, President and Co-founder, Jaytech · April 2026 · research/13-multimodal-sources/ai-for-the-c-suite/2026-04-14-dr-sam-zolfagharian-how-leaders-should-actually-approach-ai-.md

Conference Board “Governing AI” (April 2026): S&P 500 Disclosure Trajectory + Board Fluency Gap

The most authoritative independent dataset on board AI governance — drawn from public S&P 500 annual report filings plus a survey of 130 executives from U.S. public companies. Source credibility: HIGH (public filings analysis), MEDIUM-HIGH (executive survey). Non-commercial institution.

  • 12% → 83%: S&P 500 companies disclosing AI as a material risk (2023 → 2025). The legal baseline is now nearly universal; the oversight architecture to match it is not.
  • 1.5% → 2.7%: Director AI expertise (2021 → 2025). Technology expertise tripled (20% → 51%) in the same window. AI knowledge on boards is growing an order of magnitude slower than AI risk exposure.
  • 23% of boards rated “highly fluent” in AI by their own executives; 25% rated “low or no fluency.”
  • 9% of companies “very prepared” for AI regulatory compliance; 28% in early stages — with EU AI Act enforcement (Aug 2, 2026) 12 weeks out at time of publication.
  • Top board AI risk: cybersecurity/data breaches (58%), ahead of privacy (33%) and legal liability (27%).
  • 75% expect large-scale employment disruption from AI within three years; 80% expect productivity gains — boards that cannot articulate which scenario applies to their company have a governance gap.

The 12%→83% disclosure surge is the most useful boardroom data point in the corpus: it means every company in that 83% has publicly acknowledged AI as a material risk, creating an implicit standard of care for oversight.

Source: research/04-consulting-firms/conference-board-governing-ai-2026.md — n=130 executives + S&P 500 disclosure analysis, April 22, 2026. Distinct from [[bcg-split-decisions-ceos-boards-ai-2026]] (CEO-board friction, survey only) and [[bcg-five-things-boards-ai-2026]] (prescriptive framework, no survey data).

BCG AI Radar 2026 — CEO Ownership Shift and the 50% Job-Dependence Signal (Jan 2026, n=2,360)

The CEO ownership data from BCG AI Radar 2026 is the most directly relevant board-level governance finding in the January 2026 corpus.

  • 72% of CEOs now identify as the primary AI decision maker — double the share from one year prior. This consolidation of AI accountability at the CEO level means board AI oversight conversations are no longer filtered through CIO or CDO intermediaries; the CEO is directly answerable to the board on AI outcomes.
  • 50% of CEOs believe their job depends on successful AI implementation. Boards that have not explicitly aligned on what “success” means will create performance evaluation ambiguity that the BCG data identifies as a material governance risk.
  • Three CEO archetypes (Trailblazers ~15%, Pragmatists ~70%, Followers ~15%) emerge in the data. Boards should know which archetype their CEO represents: Trailblazers (8+ hours/week personal AI upskilling, 60% of AI budget to workforce development) operate very differently from Followers (cautious, 24% to workforce). A board pressuring a Follower-archetype CEO to move at Trailblazer speed without providing the resource commitments that enable it is creating execution risk.
  • 94% of CEOs plan to maintain or increase AI investment regardless of short-term returns. The board’s oversight obligation has shifted from “should we invest?” to “are we investing in the right sequence with the right governance?”

Source: research/04-consulting-firms/bcg-ai-radar-2026-ceo-mandate.md — BCG AI Radar 2026, n=2,360 (640 CEOs), 16 markets, Jan 15, 2026. MEDIUM-HIGH; BCG advisory conflict.

BCG Split Decisions Survey (May 2026): Primary data on CEO-board pace friction

The first primary survey in the corpus that matches CEO and board member respondents on the same AI governance questions. Key findings:

  • 61% of CEOs say their boards are rushing AI transformation — boards pushing for faster implementation than CEOs believe is operationally sound
  • 75% of board members rate their own AI knowledge on par with or ahead of peers; ~40% of CEOs say those same boards lack informed views on AI reshaping growth strategy — the confidence inversion is the core governance problem
  • ~33% of CEOs believe boards overestimate AI’s capacity to replace human roles — the mechanism behind governance pressure to cut headcount prematurely
  • 35% vs. 27% gap in how CEOs and board members assess the weight of AI ROI in CEO performance evaluation — when accountability is underweighted, incentives misalign between deployment pace and outcome discipline
  • 80% of both groups agree future board candidates should demonstrate measurable AI understanding — a near-unanimous mandate for making AI fluency a formal board qualification criterion

Source: research/04-consulting-firms/bcg-split-decisions-ceos-boards-ai-2026.md — n=625 (351 CEOs + 274 board members), $100M+ revenue companies, global, May 2026. Distinct from [[bcg-five-things-boards-ai-2026]] (prescriptive framework, no survey data) and [[bcg-ai-radar-2026]] (CEO-only respondents, no board member matching).

What directors now read: the 2026 consulting-firm manifesto wave

  • Forrester “The AI CIO Governs Outcomes, Not Tools” (Apr 2026) — Forrester’s IT-leadership framework for the 2030 operating model. CIOs who govern AI by output accountability (cost per outcome, time-to-value, trust metrics) rather than tool deployment outperform peers across key performance indicators. Key board-level data: 71% of CIOs report pressure to prove AI ROI within 12 months; only 29% have a formal outcome measurement framework. Directly actionable for boards asking “what should the CIO be measured on?” Source: research/04-consulting-firms/forrester-ai-cio-outcome-governance-2026.md
  • BCG “Five Things Boards Need to Get Right with AI” (Feb 24, 2026) — de Bellefonds and Lukic’s dedicated director-level framework. Opens with the single most-cited board-readiness data point: at a recent gathering of corporate directors, when asked who felt ready to oversee AI, no one raised a hand. Five items — pace and priorities, strategic freedom via the one-way doors test, deploy–reshape–invent portfolio, remuneration-committee-as-transformation-lever, measured communications + SEC AI-disclosure discipline — plus a sixth behavioral prompt (get in the sandbox and play). The remuneration-committee framing is the single most underused governance lever for mid-market boards: if compensation still rewards the legacy model, AI transformation stalls regardless of budget. Source: research/04-consulting-firms/bcg-five-things-boards-ai-2026.md
  • McKinsey AI Transformation Manifesto (Apr 2026) — QuantumBlack’s 12-theme framework derived from “hundreds of large-scale tech and AI transformations.” Headline number: across 20 AI-leading companies, the average transformation delivered 20% EBITDA uplift, broke even in 1–2 years, and generated $3 of incremental EBITDA per $1 invested. These companies concentrated on 1–3 business domains, not broad use-case portfolios. Theme 10 contains a rare McKinsey caution: “the excitement for agentic AI may be getting ahead of companies’ ability to manage the more complex risks.” Source: research/04-consulting-firms/mckinsey-ai-transformation-manifesto-2026.md
  • IBM IBV CEO Study 2026 (May 2026) — n=2,000 CEOs, Oxford Economics partnership, Feb–Apr 2026 fieldwork. The headline board-level data point: 76% of organizations now have a Chief AI Officer, up from 26% one year prior — the largest documented CAIO adoption surge in a single survey cycle. 83% of CEOs say AI sovereignty is essential to strategy; 64% are comfortable making major strategic decisions from AI input; CEOs project 48% of operational decisions will be autonomous by 2030. The 4x redesign multiplier (organizations redesigning five core areas are 4x more likely to achieve objectives) is the board-appropriate framing for why structural transformation outperforms use-case accumulation. Source: research/04-consulting-firms/ibm-ibv-ceo-study-2026.md
  • IBM IBV Agentic AI Governance Playbook (Apr 2026) — “Governance by design” framework; six operational clarity elements (ownership, authority, decision making, control, boundaries, responsibilities) and five pre-deployment tradeoffs (speed vs. control, innovation vs. predictability, accountability ownership, control evolution, automation and trust) that boards can use as a pre-deployment review checklist for any agent program. Source: research/04-consulting-firms/ibm-ibv-agentic-ai-governance-playbook-2026.md
  • Anthropic “Trustworthy Agents in Practice” (Apr 9, 2026) — Decomposes every agent into four controllable components (model, harness, tools, environment), each with a different owner and control set. Five principles — human control, value alignment, secure interactions, transparency, privacy — map to NIST AI RMF. Explicit security position: “No single line of defense is enough to guarantee protection” against prompt injection. Useful for director questions about what the company is actually buying when it buys an agent. Source: research/06-security-frontier/anthropic-trustworthy-agents-in-practice-2026.md
  • IBM IBV 5 Trends for 2026 (Dec 2025, n=1,028 C-suite + 8,500 consumers/employees) — Puts AI sovereignty on the 2026 board agenda: 93% of executives say they must factor sovereignty (control of AI systems, data, and infrastructure across providers and locations) into their 2026 strategy. 73% say physical data location matters increasingly; 58% already experienced supply-chain choke-point disruption in 2025; 75% of chip-buying firms call semiconductor-vendor concentration a major strategic challenge. Second board-ready finding: 95% of executives say consumer trust in AI will define new-product success, and two-thirds of consumers would switch brands over concealed AI involvement. Directors asking “what is the new AI risk we own?” get a concrete, quantified answer here. Source: research/04-consulting-firms/ibm-ibv-5-trends-2026.md

Morningstar Sustainalytics Proxy Voting Analysis (May 2025)

First proxy voting dataset in the corpus to quantify institutional investor support for AI oversight resolutions across named asset managers. Source credibility: MEDIUM-HIGH. TIER 2 (Morningstar Sustainalytics commercial interest in stewardship research; underlying data drawn from public proxy voting records; 35 asset managers, 15 resolutions, data as of May 7, 2025).

  • 30% average adjusted support for AI oversight resolutions in the 2024 proxy year — nearly double the 16% average for 400 general E&S resolutions. Investor demand for AI governance is structurally higher than for other ESG topics.
  • 7 of 15 resolutions crossed the 30% “significant” threshold, averaging 41% adjusted support. At this level, boards typically feel compelled to respond with either formal governance commitments or direct shareholder engagement.
  • 0 of 35 asset managers argued AI oversight was not financially material — a unanimous finding across both US and European investors. The debate between BlackRock (7% support) and European managers (77–100% support) is about whether existing company policies are adequate, not whether boards should govern AI at all.
  • US-European split by voting rationale, not materiality: US managers voting against cited “company already has adequate policies”; European managers concluded those policies are insufficient. For boards: if current disclosure does not satisfy Fidelity or MFS (both 70%+ support), a meaningful opposition bloc already exists in the shareholder base.
  • The four highest-support resolutions (≥45%) named specific harms — misinformation/disinformation and AI-driven targeted advertising at Alphabet and Meta. Specificity of demand doubles investor support relative to general governance requests.

Source: research/04-consulting-firms/morningstar-proxy-voting-ai-oversight-2025.md — Morningstar Sustainalytics Stewardship Research, Linsey Stewart CFA + River Meng, May 2025; 15 resolutions at U.S. companies; 35 asset managers (20 US, 15 European); 2024 proxy year + first 9 months of 2025. Distinct from [[iss-stoxx-board-ai-governance-gap-2026]] (company-side disclosures, not investor-side demand) and [[conference-board-governing-ai-2026]] (S&P 500 executive survey).

ISS STOXX “Mind the Governance Gap” (March 2026): Russell 3000 Public Disclosure Analysis

The only dataset in the corpus drawn from public governance filings across the full Russell 3000 (not a self-report survey), covering 3,048 U.S. companies. Source credibility: HIGH (public SEC/proxy filing analysis). TIER 1.

  • 8% of 3,048 companies disclose formal board-level AI oversight — the baseline for what “governed” actually means in 2026
  • 16% have at least one AI-skilled director; only 4% have two or more — most boards with any AI expertise are one resignation away from having none
  • 9% have formal AI policies covering development, deployment, and monitoring
  • AI governance concentrates in five sectors (Industrials, IT, Consumer Discretionary, Financials, Health Care) — 75% of all disclosed oversight, 83% of AI-skilled directors; Energy, Utilities, Materials nearly absent despite operational AI exposure
  • Skills-policy mismatch: 16% AI-skilled directors vs. 9% with policies — individual expertise without institutional guardrails is the most common failure mode
  • ISS STOXX’s Governance QualityScore feeds directly into proxy voting recommendations, making the 8% figure a competitive governance threshold, not just a benchmark

The 8% figure is the sharper companion to Conference Board’s 83% risk-disclosure stat: the gap between having told shareholders AI is a material risk (83%) and having demonstrated a governance structure to manage that risk (8%) is the core liability exposure for 2026.

Source: research/04-consulting-firms/iss-stoxx-board-ai-governance-gap-2026.md — public disclosure analysis, 3,048 Russell 3000 + S&P 500 companies, ISS Governance QualityScore January 2026 vintage. Distinct from [[conference-board-governing-ai-2026]] (executive survey + S&P 500 only) and [[bcg-split-decisions-ceos-boards-ai-2026]] (CEO-board matched survey).

EY CEO Outlook 2026 (Wave 2, May 2026): Financial Accountability Gap at CEO Level

The first quarterly CEO survey in the corpus that explicitly measures financial reporting linkage, not just investment intentions. Source credibility: MEDIUM-HIGH (FT Longitude independent fieldwork; EY-Parthenon advisory commercial interest). TIER 1.

  • Only 11% of 1,200 global CEOs tie AI impact to financial reporting reviewed regularly by senior management — despite 80% increasing AI investment
  • 82% of FS CEOs at or above expectations for AI ROI vs. only 20% all-sector — financial services boards with governance frameworks are the reference class for what “ahead” looks like
  • 48% of CEOs pursuing M&A specifically for AI capabilities — organic capability development alone is failing at scale
  • AI investment commitment held at 80% through a geopolitical shock (risk citations doubled 28%→56%) — first data point on AI investment resilience at CEO level

Source: research/04-consulting-firms/ey-ceo-outlook-2026.md — n=1,200 CEOs, 21 countries, FT Longitude independent fieldwork, May 2026. Distinct from [[bcg-ai-radar-2026]] (CEO strategy posture) and [[bcg-split-decisions-ceos-boards-ai-2026]] (CEO-board friction).

Supporting research

Protiviti / BoardProspects Global Board Governance Survey 2026 — The Board Cadence ROI Correlation (n=772, Q4 2025)

The only dataset in the 2026 corpus that directly quantifies the ROI differential between boards that discuss AI at every meeting and those that do not. Source credibility: MEDIUM-HIGH (Protiviti consulting advisory interest; no AI vendor commercial interest; third annual series; large sample). TIER 1.

  • Only 26% of corporate boards make AI a standing agenda item at every meeting — despite universal agreement that AI is a strategic priority.
  • 63% of high-ROI organizations include AI on the board agenda at every meeting. Only 13% of low-ROI organizations do the same. The 50-point gap is the strongest board-governance ROI correlation in the corpus.
  • 95% of high-ROI organizations are confident in their ability to integrate AI into operations vs. 33% of low-ROI organizations — a 62-point integration-confidence gap.
  • 93% of high-ROI organizations express confidence in their responsible AI strategy vs. 42% of low-ROI organizations — a 51-point governance-confidence gap.
  • AI-mature boards have shifted strategic focus beyond efficiency/cost reduction to customer experience, innovation, competitive positioning, and enterprise-wide scale — the areas where AI creates durable advantage, not parity.
  • The mechanism: regular board engagement creates accountability discipline — better investment decisions, earlier course correction, more rigorous demand for evidence of value. Boards that discuss AI episodically approve investments episodically and measure results episodically.

The 63/13 cadence finding lands alongside ISS STOXX’s 8% formal-oversight figure and Conference Board’s 23% board-fluency figure — three different measurement angles converging on the same structural conclusion: most boards approve AI investment without the governance infrastructure to oversee it.

Source: research/04-consulting-firms/protiviti-boardprospects-global-board-governance-survey-2026.md — n=772 board members and C-suite executives, Q4 2025 fieldwork, March 18, 2026.

Grant Thornton 2026 AI Impact Survey — The Investment/Governance Disconnect (n=950, Feb–Mar 2026)

The sharpest board-governance dataset in the corpus for mid-market companies (Grant Thornton’s client base skews 200–2,000 employees; Big 4 adjacent, no AI product commercial interest). MEDIUM-HIGH credibility. TIER 1.

  • 74% of boards approved major AI investments. 48% have not set AI governance expectations. The spend is flowing; the oversight is not following.
  • 54% have not integrated AI risk into ongoing board oversight — despite 12%→83% of S&P 500 companies (Conference Board 2026) now disclosing AI as a material risk in annual filings. Disclosure without oversight is the liability posture.
  • 78% of executives lack confidence they could pass an independent AI governance audit within 90 days. For piloting organizations, only 7% are “very confident.” For fully integrated organizations, that figure rises to 74% — governance accumulates through operational decisions made during integration, not through policy documents approved in the board room.
  • Agentic AI governance is the most urgent board gap: 73% of these organizations are already giving autonomous agents access to enterprise data and processes. Only 20% have tested an AI incident response plan. Boards approving agentic AI deployments without incident response infrastructure are approving deployments without a risk management floor.
  • The COO signal boards should track: 54% of COOs cite regulatory and compliance uncertainty as their primary agentic AI concern, vs. 20% of CIOs/CTOs. The person closest to operational consequences is five times more concerned than the person presenting the technology investment. When those two numbers diverge, the board is receiving a biased risk picture.

Source: research/04-consulting-firms/grant-thornton-ai-impact-survey-2026.md · Feb–Mar 2026 fieldwork, April 2026 published · MEDIUM-HIGH · TIER 1

EY CEO Outlook 2026 — The Board-Level Accountability Gap (n=1,200, FT Longitude, Jan + May 2026)

  • Only 11% of global CEOs say AI impact is linked to financial reporting and reviewed regularly by senior management — despite 80% committing to increased investment. The board governance implication: if CEO accountability for AI is not in financial reporting, board oversight of AI is almost certainly performative. The 11% cohort is the pool from which AI value leaders are drawn.
  • AI investment proved recession-resilient: Between September 2025 and April 2026, geopolitical risk citations by CEOs doubled (28%→56%), but AI investment commitment held at 80%. Boards modeling AI budget scenarios as discretionary are working from an outdated risk model.
  • Financial services boards are leading on governance: 90% of FS CEOs (n=240) report board-level AI governance frameworks, vs. ~80% in non-financial sectors. The FS board posture is 2–3 years ahead of non-regulated industries on both governance infrastructure and outcomes.

Source: research/04-consulting-firms/ey-ceo-outlook-2026.md · Jan + May 2026 · TIER 1 · MEDIUM-HIGH

What this means for mid-market buyers

  • A 200-500 person company needs a 5-8 slide board briefing, not a 50-slide deck. Directors want narrative, not noise: what are you doing with AI, what is the risk, what should you do next.
  • Structure quarterly board time for AI discussion. 62% of directors now set aside full-board agenda time for AI (NACD 2025 Board Practices Survey). Companies that structure this time capture the oversight benefit; those that treat it as a CTO monologue waste it.
  • Assign a named executive sponsor. Projects with sustained CEO-level sponsorship achieve 68% success rates versus 11% for those that lose executive engagement within 6 months (Pertama Partners, 2026).

Dataiku Global AI Confessions — CEO Edition 2026 (n=900 CEOs, Harris Poll, Feb–Mar 2026)

  • 80% of CEOs say their job is at risk by end of 2026 without delivering measurable AI results — up from 74% who said the same a year ago. The accountability horizon has shortened from “within two years” to “within months.”
  • 72% of U.S. CEOs report board pressure for measurable AI outcomes, up from 61% in 2025. Boards have moved from asking about AI strategy to demanding proof of return.
  • 81% of U.S. CEOs believe a fellow CEO will be ousted due to a failed AI strategy in 2026 — these executives are scanning for cautionary examples and adjusting behavior accordingly.
  • 62% globally are challenged by their boards on AI outcomes, and 83% plan full production AI agent deployment this year — creating a high-stakes overlap where boards will be evaluating results from systems not yet proven at scale.
  • Confidence in deploying AI agents at scale dropped from 41% to 31% year-over-year — the most significant confidence decline in the 2026 CEO research corpus.

Source: research/01-ai-native-landscape/dataiku-global-ai-confessions-ceo-2026.md · Feb–Mar 2026 · MEDIUM · TIER 1

See also

  • Agentic AI Governance — the governance architecture boards are now expected to oversee
  • AI Washing and Enforcement — SEC/FTC enforcement exposure that boards are accountable for
  • AI Budget and CFO Decision-Making — the three-year cost arc that turns a board discussion into a budget commitment
  • Model Risk Management — SR 11-7 framework directors should recognize in regulated industries
  • AI Sovereignty — 93%-of-executives board-level theme for 2026: data residency, compute-vendor concentration, model portability
  • SEC AI Disclosure — IAC three-requirement framework, Reg S-K item mapping, and enforcement phases; the securities-law dimension of board AI governance

Grant Thornton 2026 AI Impact Survey (n=950, Feb–Mar 2026)

  • 75% of boards approved major AI investments; only 52% set clear AI governance expectations. Capital allocation is running 23 points ahead of accountability architecture — the board approved the spend without requiring the oversight infrastructure.
  • Only 54% of boards have integrated AI risk into their ongoing oversight process. The investment is approved; the monitoring is not structured.
  • Organizations where boards have set AI governance expectations are 10x more likely to pass an independent audit and 4x more likely to report AI-driven revenue growth. Governance is the mechanism through which board investment converts to financial returns.
  • “Most governance models weren’t designed for AI.” — Tom Puthiyamadam, Grant Thornton. The urgency: the August 2, 2026 EU AI Act prohibited-use deadline and growing SEC/FTC enforcement means “wait and see” on governance is no longer a defensible board posture.

Source: research/07-adoption-challenges/grant-thornton-ai-impact-survey-2026.md · Grant Thornton n=950 Feb–Mar 2026 · HIGH · TIER 1

McKinsey State of Organizations 2026 — Governance Accountability Gap (n=10,018, Jun–Sep 2025)

The largest organizational survey of 2026 names governance failure — not AI capability — as the primary barrier to bottom-line impact.

  • 88% of organizations deploy AI; 81% report no meaningful bottom-line impact. The gap is organizational design, not technology readiness.
  • 1 in 6 leaders reports no clear C-suite owner of AI. Boards that have not required a named accountable executive are funding diffuse responsibility, not transformation.
  • Only 14% of leaders consistently champion AI adoption with a clear strategy. Among AI Pioneers — the 23% of organizations realizing results — 90% of leaders actively champion adoption. The performance gap traces directly to board expectations of leadership behavior.
  • High performers are 3x more likely to have senior leaders demonstrating ownership and commitment to AI. Board-level tone setting is a measurable performance variable, not a governance formality.
  • McKinsey’s prescriptive ratio: $5 in people for every $1 spent on AI technology. Organizations that follow this ratio are 4x more likely to sustain top-tier financial performance over a decade. Boards approving technology budgets without a parallel people investment mandate are under-capitalizing the transformation.

Source: research/04-consulting-firms/mckinsey-state-of-organizations-2026.md · McKinsey People & Organizational Performance, n=10,018, Jun–Sep 2025 · MEDIUM-HIGH · TIER 2

Oliver Wyman Forum / NYSE — CEO Agenda 2026: Board Oversight & CEO AI Accountability (n=415, $13T market cap, April 2026)

CEO-level board engagement data from the most capital-weighted sample available.

  • 96% of CEOs report increased board involvement in their companies — strategy/governance (61%), executive performance/succession (35%), risk management (34%)
  • 50% of CEO planning time now dedicated to horizons under one year — up from 43% in 2025; boards are pulling CEOs into short-term governance at the expense of multi-year strategy
  • 11% of CEOs were replaced in 2025 (Spencer Stuart data) — tenure pressure is real; AI delivery expectations are now a CEO performance criterion
  • AI ROI confidence fell year-over-year: only 27% report ROI meeting expectations (down from 38%); 12% are “AI ROI leaders” (down from 17%) — boards are asking, returns are not materializing at the rate presentations implied
  • The board accountability pattern: boards approved capital (75% per Grant Thornton), set governance expectations only 52% of the time, and are now measuring CEO performance against AI returns that 73% of CEOs cannot yet articulate

Source: research/04-consulting-firms/oliver-wyman-ceo-agenda-ai-workforce-2026.md · Oliver Wyman Forum / NYSE, n=415, April 2026 · HIGH · TIER 1

Information Integrity Risk: What Audit Committees Need to Ask (Gartner Q1 2026)

Gartner’s Q1 2026 Emerging Risk Report (n=337 ERM/audit/risk executives) put information integrity risk at #1 — the risk that AI-enabled decisions run on untrustworthy or opaque data with no audit trail. This is the framing audit committees need to use when asking AI governance questions.

  • First time an AI-native risk has been #1 in Gartner’s Quarterly Emerging Risk series — displacing geopolitics and cybersecurity
  • Board question to ask: which AI systems in this organization are influencing decisions, and do those systems produce legible audit trails showing what data drove which conclusions?
  • Second new risk: AI workforce preparedness — audit committees should ask whether any named executive is accountable for ensuring AI-informed decisions meet the same evidentiary standard as human-informed decisions
  • Risk concerns identified by Gartner: pace of AI innovation outrunning governance, regulatory transparency uncertainty, malicious use of AI-generated information, and leadership communication failures about AI-derived conclusions
  • Recommended response: inventory AI deployments that influence decisions (not just generate content), assess data provenance and explainability, assign accountable executive ownership

Source: research/05-analyst-firms/gartner-information-integrity-risk-q1-2026.md · Gartner Quarterly Emerging Risk Report · Q1 2026 · HIGH · TIER 1

MIT CISR AI-Savvy Boards 2025 — The 14.7-Point ROE Spread (n=2,800, Dec 2025)

The most quantitatively precise board AI governance dataset in the corpus. Primary research analyzing 2,800 publicly traded companies with $1B+ revenue, measuring return on equity and market cap against board AI expertise. Source credibility: HIGH. Independent academic research, no vendor sponsorship. TIER 1 (Q4 2025).

  • Only 26% of large U.S. company boards qualify as “AI-savvy” under MIT CISR’s updated criteria, which requires 3+ directors with hands-on experience in generative AI, AI agents, robotics, and related technologies. 72% still meet only the 2019 “digitally savvy” bar — which was designed for cloud and mobile, not GenAI.
  • AI-savvy boards: +10.9 pp ROE above industry average. Non-savvy boards: -3.8 pp below. The 14.7-point spread is the strongest financial performance signal associated with board composition in the 2026 corpus.
  • Market cap premium: AI-savvy companies average $15.5B above their industry peers; non-savvy companies average $5.4B below — a $20.9B structural valuation gap.
  • Industry breakdown by board AI-savviness: Information services (68%), professional services (52%), finance/insurance (31%), health care (8%), retail-automotive (11%), construction (6%), mining (4%).
  • Health care at 8% is the starkest mismatch — one of the highest AI deployment sectors governed by boards that almost universally lack hands-on AI expertise.
  • This continues the pattern MIT CISR established in 2019 (digitally savvy boards outperformed on revenue growth +48%, profit margins +16%) — the dynamic scales and strengthens as the technology becomes more consequential.
  • Definition of AI savvy: 3+ directors with hands-on experience in cutting-edge AI and emerging technologies (updated from 3+ “digital” directors in 2019).

The companion board-function framework: successful AI-savvy boards operate along three axes — Strategy (evaluating AI opportunities and threats), Defense (managing cyber risk and regulatory compliance), and Oversight (tracking value creation and ensuring responsible data use). Non-savvy boards conflate all three into a single CIO briefing.

Source: research/04-consulting-firms/mit-cisr-ai-savvy-boards-superior-performance-2025.md · Peter Weill, Stephanie Woerner, Jennifer Banner · MIT Sloan Management Review, Winter 2026 · n=2,800 companies with $1B+ revenue · Dec 8, 2025. Distinct from [[mit-cisr-minimum-viable-governance-2026]] (governance program design) and [[iss-stoxx-board-ai-governance-gap-2026]] (public disclosure analysis, Russell 3000).

MIT CISR Minimum Viable Governance 2026 — The Board Oversight Trap

MIT CISR Research Briefing Vol. XXVI, No. 3 (March 19, 2026). One-year case study plus 17 executive interviews.

The FinCo case study is directly relevant to board AI strategy: a board-sponsored AI governance program, fully resourced and endorsed at the highest level, produced more shadow AI than it prevented. The mechanism is a board strategy pattern — mandate comprehensive governance without accounting for the cost of that governance’s latency, and the organization fills the vacuum with unmanaged tools.

MIT CISR’s four characteristics define what the board should hold management accountable for delivering:

  • Structurally Agile: tiered review matched to risk, not uniform committee approval for all initiatives
  • Trustworthy by Design: oversight embedded in platforms, not gated at access points
  • Integrated End-to-End: risk and compliance embedded in delivery teams from day one
  • Opportunity-Sensitive: time-to-decision tracked as a board metric alongside risk incidents

Board members asking about AI governance should request data on time-to-decision for low- and high-risk AI initiatives, not just incident counts. A governance program that blocks a low-risk prototype for six months while tracking zero risk incidents is failing on the dimension that drives shadow AI.

Source: research/06-security-frontier/mit-cisr-minimum-viable-governance-2026.md · MIT CISR, March 2026 · HIGH · TIER 1