---
title: "AI Strategy"
date: 2026-07-12
description: Approving good AI projects is not a strategy, and the Board's move is from accumulating pilots to a strategy it owns.
author: Mario Thomas
canonical: https://mariothomas.com/briefings/ai-strategy/
---

## Start here

Two short reads before going deeper: what AI strategy actually is, and the order that makes this briefing most useful.

### Strategy Is Not a Stack of Business Cases

McKinsey's 2025 Superagency in the Workplace report found **92%** of
organisations planning to increase AI investment, yet only **1%**
describing their adoption as mature. The Boards I meet are not short of
AI activity. They are short of a way to make it add up, and the gap
usually traces back to one habit: approving AI project by project,
business case by business case, and calling the accumulation a strategy.

A business case is a tool of isolation. It asks whether one initiative
justifies its cost; it cannot ask whether that initiative reinforces or
undermines the others, builds capability the next one needs, or leaves
the organisation with three incompatible approaches to data governance.
S&P Global's 2025 analysis found **42%** of organisations scrapping most
of their AI initiatives, up from 17% the year before. In my reading those
were rarely technical failures. Many of the pilots worked; the strategy
they were supposed to add up to never existed.

The position these articles take is Richard Rumelt's: a strategy is a
diagnosis of the challenge, a guiding policy for addressing it, and
coherent actions that reinforce one another. Business cases skip the
first two and jump straight to isolated actions. The core sequence here
works through all three, treating the Six Board Concerns model as an
interconnected diagnosis, the Complete AI Adoption Framework (the lineage
that became Remake) as guiding policy, and a sequenced playbook as the
actions. Along the way it takes two realities governance tends to resist,
multi-speed adoption and shadow AI, and treats them as assets rather than
failures.

A director who reads this briefing should be able to make one judgement
with confidence: whether the organisation currently has an AI strategy
or a collection of approvals, and what the next AI paper that reaches
the Board is actually asking it to govern.

### Trap, Diagnosis, Policy, Actions: The Reading Order

Start with [the articles](#core-reading). The core sequence is numbered
and written to be read in order, because it follows Rumelt's structure.
It opens by naming the trap: mistaking accumulated business cases for
strategy. It then diagnoses the real challenge through the Six Board
Concerns, not as a checklist but as an interconnected system. Next it
sets out the Complete AI Adoption Framework as guiding policy for
orchestrating adoption that moves at different speeds. It closes by
turning policy into a sequenced playbook, from a Day 1 AI amnesty through
to the scaling decisions of Quarter 4.

The supporting pieces supply context and correctives. The Complete AI
Adoption Framework article is the prequel, explaining how the AI Stages
of Adoption model, the Five Pillars diagnostic, and the Well-Advised
principle fit together before the series puts them to work. The
diagnosis builds on the earlier piece on Board priorities for AI
governance, where the Six Board Concerns were first set out, and the
playbook's Day 1 action rests on the after-the-amnesty guide, its
implementation blueprint for turning disclosed shadow AI into governed
capability. The year-in-review and the five forces briefing set the
context a director walks into in 2026. The maturity mirage and the
return of traditional AI are the correctives: one tests whether the
progress being reported to the Board is real, the other whether the
organisation is deploying the right kind of AI at all. The 2024 piece on
AI beyond the generative hype is the background to that second test, and
the piece on why not everything needs AI asks the question that comes
before either: whether the work needs doing at all.

Then go to [Remake](#remake-assets) for the mechanisms beneath the
thinking. They are the apparatus the articles keep reaching for, and
each links through to its Library entry. Finish with
[the questions](#faqs), which compress the briefing into the questions I
am most often asked, and the references for the primary evidence behind
the statistics.

With 15 minutes to spare, read the opening article and the questions;
the rest of the core sequence will keep until the flight home.

## Core reading

A Rumelt-shaped sequence: name the trap, diagnose the challenge, set the guiding policy, then act coherently. The supporting pieces supply context and correctives.

1. [From AI Pilots and Projects to AI Strategy: Avoiding the Business Case Trap](https://mariothomas.com/blog/ai-strategy-business-case-trap/) (10 minute read, 5 October 2025): Approving AI business cases at record pace is not a strategy: 92% are investing more, 1% have reached maturity. The gap is coherence.
2. [AI's Interconnected Challenge: Diagnosing the Six Concerns of the Board](https://mariothomas.com/blog/ai-strategy-diagnosis/) (12 minute read, 12 October 2025): The Board's six concerns demand simultaneous orchestration and receive sequential, project-level attention. Treating them as one diagnostic lens is where AI governance starts.
3. [Orchestrating Multi-Speed AI: The Complete AI Framework as Guiding Policy](https://mariothomas.com/blog/ai-strategy-guiding-policy/) (12 minute read, 19 October 2025): Most organisations use AI; few have redesigned the work around it. The Complete AI Framework is the guiding policy turning diagnosis into action.
4. [Completing the AI Strategy Journey: From Policy to Practice Through Coherent Actions](https://mariothomas.com/blog/ai-strategy-coherent-actions/) (14 minute read, 26 October 2025): 69% of Boards discuss AI regularly; 33% feel equipped to oversee it. Closing the execution gap means policy becoming actions that compound rather than fragment.

## Further reading

- [Not Everything Needs AI: The Questions That Come Before the Decision](https://mariothomas.com/blog/not-everything-needs-ai/) (8 minute read, 12 July 2026): Boards keep being told to remake the business around AI. The real question is not which tool, but whether the work needs doing at all. Podcast edition: 8 minute listen.
- [A Complete AI Adoption Framework: AISA, Five Pillars, and Well-Advised](https://mariothomas.com/blog/complete-ai-framework/) (15 minute read, 1 June 2025): The AI Stages of Adoption, the Five Pillars, and Well-Advised are one framework: how they integrate to govern multi-speed AI adoption across the organisation.
- [Return-to-Work Briefing: Five Forces Reshaping the Board AI Agenda in 2026](https://mariothomas.com/blog/five-forces-board-ai-agenda/) (10 minute read, 4 January 2026): Five forces shape the Board's AI agenda in 2026, led by AI embedding into the enterprise faster than it can be governed. None is distant. Podcast edition: 15 minute listen.
- [The Year AI Grew Up: Five Inflections That Changed the Strategic Calculus in 2025](https://mariothomas.com/blog/the-year-ai-grew-up/) (14 minute read, 30 December 2025): In 2025 Boards stopped asking what AI could do and started treating it as strategic infrastructure investment. Five connected inflections drove that shift. Podcast edition: 19 minute listen.
- [The AI Maturity Mirage: Diagnosing the Gap Between Investment and Readiness](https://mariothomas.com/blog/ai-maturity-mirage/) (11 minute read, 7 December 2025): Boards overestimate AI maturity by counting tools and pilots rather than capability. Three patterns create the illusion, and each can be diagnosed before it misleads. Podcast edition: 16 minute listen.
- [The Return of Traditional AI: Organisations Are Rethinking Their LLM-First Strategies](https://mariothomas.com/blog/return-of-traditional-ai/) (8 minute read, 21 December 2025): Forty-two percent of companies abandoned most of their AI initiatives this year, often because generative AI was applied to problems traditional methods solve better. Podcast edition: 13 minute listen.
- [Beyond the hype: Unlocking the true potential of AI in business](https://mariothomas.com/blog/beyond-the-hype-generative-ai/) (10 minute read, 19 June 2024): Generative AI is one branch of artificial intelligence, not the whole story. Chasing only the hype misses the kinds of AI already creating value.
- [After the AI Amnesty: Practical Steps to Operationalise Discovered Shadow AI](https://mariothomas.com/blog/shadow-ai-amnesty-next-steps/) (12 minute read, 28 September 2025): After the amnesty, speed matters: employees who disclosed expect enablement, not restriction. A roadmap for turning discovered shadow AI into governed capability.
- [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.

## 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: Five Pillars of AI Capability**. The five capability domains of an AI capability model that cut across every level of maturity: Governance and Accountability, Technical Infrastructure, Operational Excellence, Value Realisation and Lifecycle Management, and People, Culture and Adoption. [Remake Library](https://mariothomas.com/remake/library/five-pillars/)
- **Methodology: AI Amnesty**. The time-boxed programme that brings shadow AI under governance: a declaration window, typically 30 to 45 days, in which employees disclose all AI tool usage without fear of punishment, followed by the post-amnesty roadmap of rapid triage, governance guardrails, pilot launch, and ongoing operations. [Remake Library](https://mariothomas.com/remake/library/#ai-amnesty)
- **Principle: Well-Advised**. The framework of five strategic priorities, Innovation, Customer Value, Operational Excellence, Responsible Transformation, and Revenue, used to ensure AI investments create balanced value rather than narrow cost reduction. [Remake Library](https://mariothomas.com/remake/library/well-advised/)

## Questions

The questions directors most often put to me on this subject, answered from the articles in this briefing.

### We subject every AI initiative to a rigorous business case. Why is that not enough?

Because a business case is a tool of isolation: it evaluates one initiative against its own costs and returns, and cannot see whether initiatives reinforce or undermine each other. S&P Global's 2025 analysis found 42% of organisations scrapped most of their AI initiatives; in my reading they were rarely technical failures. [The opening article](/blog/ai-strategy-business-case-trap/) sets out why accumulating approvals fragments capability rather than compounding it.

### What does a real AI strategy actually contain?

Rumelt's three elements: a diagnosis of the challenge, a guiding policy for addressing it, and coherent actions that reinforce one another. In this briefing the diagnosis is [the Six Board Concerns as an interconnected system](/blog/ai-strategy-diagnosis/), the guiding policy is [the Complete AI Adoption Framework](/blog/ai-strategy-guiding-policy/), the lineage that became Remake, and the actions are [a sequenced playbook](/blog/ai-strategy-coherent-actions/) running from AI amnesty to scaling. A document that cannot show all three is a plan for spending, not a strategy.

### Our functions are adopting AI at completely different speeds. Should we force them into line?

No, and I would treat the variation as an asset. Marketing racing ahead and finance moving deliberately is the natural shape of AI adoption; forcing a uniform pace wastes the leaders' advantage and rushes the laggards' foundations. [The guiding policy article](/blog/ai-strategy-guiding-policy/) shows how advanced functions become learning laboratories whose governance and capabilities pull others forward, provided the Board orchestrates the speeds rather than synchronising them.

### What do we do about the AI tools our people use without approval?

Start with amnesty, not prohibition. MIT's 2025 State of AI in Business report found workers at over 90% of organisations using personal AI tools, and that ungoverned activity is evidence of where value actually lives. [The coherent actions playbook](/blog/ai-strategy-coherent-actions/) opens with a 30-day amnesty precisely because visibility has to precede governance; suppression drives the innovation deeper underground while the risks remain.

### How do we know whether the AI progress reported to us is real?

Treat tool counts and pilot wins with suspicion; they are the raw material of [the maturity mirage](/blog/ai-maturity-mirage/). Map each function against the AI Stages of Adoption, test capability balance across the Five Pillars, and watch one indicator in particular: pilots that consistently fail to scale within six months mean the organisation is still Experimenting, whatever the deck says. [The return of traditional AI](/blog/return-of-traditional-ai/) adds the companion question: are we deploying the right kind of AI for the problem at all?

### Who has the standing to say a piece of AI work should stop?

The Board, and in my experience nobody below it. The people closest to a piece of work are the least likely to conclude that it should stop, because their expertise, their team, and often their standing are bound up in it continuing. That is why the question that comes before any tool decision, how the work is done today and whether it still needs doing, is a governance question rather than a procurement one. [Not everything needs AI](/blog/not-everything-needs-ai/) argues that the most valuable answer is sometimes that the work should stop, and that a well-judged no is what makes every AI yes credible.

### What should the first year of our AI strategy look like?

Sequenced, each action building the capability the next one needs. [The coherent actions playbook](/blog/ai-strategy-coherent-actions/) opens on Day 1 with an AI amnesty, a 30-day window for disclosing the tools people already use, then stands up an AI Centre of Excellence to turn those discoveries into governed capability. Quarter 1 orchestrates the portfolio, Quarter 2 installs leading, lagging, and predictive metrics mapped to the Six Board Concerns model, and Quarters 3 and 4 scale what worked and retire what did not. Deloitte's 2025 Global Board Survey found 69% of Boards discussing AI regularly and 33% feeling equipped to oversee it; the playbook is written to close that gap.

### Which parts of our AI strategy does the Board itself own?

The parts a business case cannot supply. [The opening article](/blog/ai-strategy-business-case-trap/) asks the Board to stop being an approval gate for individual cases and become the orchestrator of the whole: owning the diagnosis of the challenge, the guiding policy that aligns initiatives with it, and the test of whether each initiative builds capability rather than merely clearing a return hurdle. [The playbook's boardroom commitment](/blog/ai-strategy-coherent-actions/) adds what only a Board can give: patient capital that prioritises learning over immediate returns, oversight that keeps pace with AI instead of the quarterly cycle, and a willingness to use the tools itself. Management runs the initiatives; the Board owns whether they add up.

## References

The primary research behind the statistics in these articles; every figure cited in the briefing carries a named source and year.

- **McKinsey & Company** (5 November 2025): [The State of AI](https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai). The adoption and workflow-redesign evidence the core sequence rests on, including the 21% who have redesigned workflows to integrate AI.
- **McKinsey, Superagency in the Workplace** (28 January 2025): [Superagency in the workplace: Empowering people to unlock AI’s full potential](https://www.mckinsey.com/capabilities/mckinsey-digital/our-insights/superagency-in-the-workplace-empowering-people-to-unlock-ais-full-potential-at-work). Source of the gap that opens the series: 92% planning increased AI investment against 1% reaching maturity.
- **Stanford HAI, 2025 AI Index Report** (April 2025): [The 2025 AI Index Report](https://hai.stanford.edu/ai-index/2025-ai-index-report). Independent benchmarks on AI adoption, investment growth, and accelerating legislative activity.
- **Deloitte, Global Board Survey** (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). Finds 69% of Boards discussing AI regularly while only 33% feel equipped to oversee AI strategy.
- **S&P Global Market Intelligence** (30 May 2025): [Generative AI experiences rapid adoption, but with mixed outcomes – Highlights from VotE: AI & Machine Learning](https://www.spglobal.com/market-intelligence/en/news-insights/research/ai-experiences-rapid-adoption-but-with-mixed-outcomes-highlights-from-vote-ai-machine-learning). Documents the jump in scrapped AI initiatives from 17% to 42% in a single year.
- **RAND Corporation** (13 August 2024): [The Root Causes of Failure for Artificial Intelligence Projects and How They Can Succeed](https://www.rand.org/pubs/research_reports/RRA2680-1.html). The 70-85% AI project failure estimates that anchor the failure-rate debate.
- **Gartner** (29 July 2024): [Gartner Predicts 30% of Generative AI Projects Will Be Abandoned After Proof of Concept By End of 2025](https://www.gartner.com/en/newsroom/press-releases/2024-07-29-gartner-predicts-30-percent-of-generative-ai-projects-will-be-abandoned-after-proof-of-concept-by-end-of-2025). The prediction that 30% of generative AI projects would be abandoned after proof of concept by end-2025.
- **BCG, The Widening AI Value Gap** (September 2025): [The Widening AI Value Gap](https://media-publications.bcg.com/The-Widening-AI-Value-Gap-October-2025.pdf). Survey of more than 1,250 firms finding only 5% achieve AI value at scale.
- **NACD, 2025 Governance Outlook** (12 December 2024): [Tuning Corporate Governance for AI Adoption](https://www.nacdonline.org/all-governance/governance-resources/governance-research/outlook-and-challenges/2025-governance-outlook/tuning-corporate-governance-for-ai-adoption/). The director-level view of AI readiness, including the 31% of organisations unprepared to deploy at scale.
- **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). Twelve practical principles for directors exercising AI oversight, updated for 2025.
- **MIT NANDA** (July 2025): [The GenAI Divide: State of AI in Business 2025](https://mlq.ai/media/quarterly_decks/v0.1_State_of_AI_in_Business_2025_Report.pdf). The State of AI in Business 2025 report: over 80% of organisations have piloted generative AI and around 5% have taken it into production, the divide the strategy articles are written against.
- **World Economic Forum** (12 August 2025): [Why AI is replacing some jobs faster than others](https://www.weforum.org/stories/2025/08/ai-jobs-replacement-data-careers/). The World Economic Forum's 2025 analysis of why AI is replacing some jobs faster than others, cited across the strategy series.

## The ideas beneath this briefing

Ideas I've named and matured writing about AI Strategy: what each one means, and where it started.

- **Business Case Trap**: The misconception that accumulating individual AI business cases, each justified in isolation, constitutes AI strategy, creating fragmentation, incompatible governance and pilots that never cohere into competitive advantage. [Read more](https://mariothomas.com/blog/ai-strategy-business-case-trap/)
- **Emergent Threat Paradox**: AI risks evolve through learning, adaptation and interaction in ways traditional risk frameworks cannot anticipate, so established controls fail against systems that continuously learn and change. [Read more](https://mariothomas.com/blog/ai-strategy-diagnosis/)
- **Multi-Speed Collision**: When functions align AI to their own objectives at different velocities, their individual successes actively undermine each other, for example marketing generating demand that supply chain AI cannot fulfil. [Read more](https://mariothomas.com/blog/ai-strategy-diagnosis/)
- **Trust Multiplier Effect**: How stakeholder confidence cascades, employee doubt breeding customer suspicion, alerting regulators, spooking investors, so lost trust in one group turns technical triumphs into organisational disasters. [Read more](https://mariothomas.com/blog/ai-strategy-diagnosis/)
- **Value Attribution Crisis**: The difficulty of measuring AI value that emerges through compound effects defying simple attribution, causing project-based evaluation to systematically undervalue transformation while overvaluing incrementalism. [Read more](https://mariothomas.com/blog/ai-strategy-diagnosis/)

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 Business Cases](https://mariothomas.com/briefings/ai-business-cases/): Business cases built for predictable payback misread AI, whose value arrives in parallel, late, and elsewhere; Boards need different instruments.
- [Operating AI](https://mariothomas.com/briefings/operating-ai/): Most AI pilots never reach production; the AI Centre of Excellence is the operating capability that turns scattered experiments into governed, scaled adoption.
- [The Great Remaking](https://mariothomas.com/briefings/the-great-remaking/): AI is restructuring how organisations think, decide, create, and deliver, and the gap between those that redesign work and those that wait compounds.