How Directors Can Start Using AI Before Their Boards Officially Adopt It
A practical four-phase system, with ten concrete tools, for directors who cannot yet upload board material into AI
In conversations with directors and governance professionals, there is growing curiosity about how artificial intelligence can be used even before organisations formally allow it. That is especially true on boards where confidential material still cannot be uploaded into AI systems, even as the pressure to stay informed, prepare well, and think more broadly keeps increasing.
That pressure is not abstract. The World Economic Forum’s Global Risks Report 2026 describes the current period as an “age of competition”, with geoeconomic confrontation now ranked as the top global risk for the year, ahead of state-based armed conflict, extreme weather, societal polarisation and misinformation. AI-related risks climb sharply over the ten-year horizon, and 68% of the experts surveyed expect a multipolar or fragmented international order over the next decade. For boards, that combination — geopolitical fragmentation, technological acceleration, and stakeholder volatility — makes continuous strategic awareness less a luxury than a basic governance requirement.
My own starting point was the same. The interest began with curiosity and gradually became a more disciplined practice: learning to use AI while keeping confidentiality firmly in mind, experimenting without exposing sensitive information, and exploring where these tools can still add value when formal adoption is not yet in place.
One conversation that has stayed with me came from the Harvard Corporate Director Certificate programme at Harvard Business School Executive Education, particularly with Professor Suraj Srinivasan, whose work sits at the intersection of corporate governance, boards, and AI-related business transformation. One idea from that experience has remained especially powerful: anyone who wants to become genuinely good at understanding AI has to keep trying it in practice, rather than treating it as a purely theoretical subject.
That mindset shaped the learning process. It encouraged a routine of experimentation: attending conferences, speaking with people who use these tools every day, exchanging ideas with younger professionals deeply immersed in the space, and trying new use cases regularly, even when the field is hard to keep up with because it moves so fast.
Another valuable environment has been the IE Generative AI Learning Community at IE University, where Mario Sánchez Moreno, Director of Liquid Learning, leads practical experimentation around generative AI and applied use cases. That kind of exposure matters because it expands the imagination beyond governance alone, including applications in training, education, design, and other creative formats that help people grasp the broader potential of these tools.
Confidentiality, however, remains the essential boundary. On boards where documentation cannot yet be uploaded, the discipline has to be strict: no explicit confidential information, no implicit clues that might allow sensitive material to be inferred, and no shortcuts simply because the tools are powerful. The point is not to bypass governance discipline, but to strengthen judgment around governance while institutional frameworks are still catching up.
What follows is a practical way of thinking about that process, structured around ten concrete tools across four phases: building continuous strategic awareness and foresight throughout the year, preparing before the meeting, challenging the issues once the documentation arrives, and making sense afterwards of how the market and stakeholders are interpreting decisions. This sits within the broader work I do through AIRIS on board-level decision quality and the responsible use of AI in governance.
Phase 1 — Continuous Strategic Awareness and Foresight
The first phase happens long before any board papers arrive. It is essentially about foresight: building the situational awareness and the future-oriented thinking that should be in place well before any specific decision reaches the board table. Directors today are expected to remain informed across an almost impossible range of issues — geopolitics, cybersecurity, AI, stakeholder expectations, infrastructure resilience, reputational risk, technological disruption, regulation — and to do so with enough depth to anticipate, not merely react. The WEF Global Risks Report 2026 is a useful reminder that several of these domains are now actively reshaping each other: geoeconomic confrontation, supply-chain fragmentation, AI risks, cyber insecurity and societal polarisation are interacting in ways that were unusual just a few years ago.
The problem is no longer access to information. It is filtering relevance from noise, and connecting that relevance to a longer time horizon than the next meeting.
Tool 1 — The Strategic Radar
The first tool in this phase is a strategic radar. With recurring tasks, saved prompts, alerts, and source-based monitoring, directors can create an ongoing flow of intelligence around the themes that matter most to their boards and sectors.
“Act as a strategic intelligence analyst for the board of a [sector] company operating in [region]. Every morning, provide a concise briefing covering the most important geopolitical, technological, regulatory, and reputational developments that could become strategically relevant over the next 3–5 years. Prioritise weak signals, second-order effects, and changes in stakeholder expectations.”
Used this way, AI is not simply aggregating information. It helps create a form of ambient strategic awareness that sharpens attention over time, and that can later feed into more deliberate foresight work.
Tool 2 — The Weak Signal Scanner
Closely linked to the radar is weak-signal detection, which is one of the most concrete forms foresight takes in practice. Many disruptions appear first as scattered developments before they become visible trends. Small changes in political discourse, investor expectations, technological development, regulatory language, or public sentiment often remain peripheral long before they become strategically material. Recent examples are easy to find: shifting export controls on semiconductors, the rapid normalisation of AI use in financial supervision, accelerating litigation around climate disclosure, and the redrawing of energy supply chains away from concentrated dependencies.
AI is particularly useful at connecting these fragments across domains that directors rarely have time to connect manually.
“Act as a foresight and strategic anticipation advisor. Identify emerging trends and weak signals related to AI governance, infrastructure resilience, cybersecurity, geopolitics, and energy transition that could materially affect European companies over the next decade.”
“Review the latest publications from the World Economic Forum, OECD, European Commission, IEA, and leading think tanks. Identify under-discussed developments and strategic risks that boards should monitor over the next 5–10 years.”
The value rarely comes from prediction in any narrow sense. It comes from expanding the range of futures considered before strategic discussions even begin.
Tool 3 — The External Scenario Builder
A third tool in this phase is the construction of external scenarios. This is foresight in its more structured form. The point here is important and worth making explicit: the scenarios built at this stage are not scenarios about the company’s strategy. They are scenarios about the wider environment in which the company operates.
That distinction matters for two reasons. First, because confidentiality boundaries do not allow internal strategy to be uploaded into general AI tools. Second, because foresight is most useful when it is genuinely external: thinking through how geopolitics, cybersecurity, sustainability, the energy transition, technological change, or regulation might evolve, regardless of any specific company. Those external scenarios can then inform internal strategic discussions later, in fully confidential settings.
“Using the WEF Global Risks Report 2026 and recent publications from the OECD, IEA, and leading think tanks, build three plausible scenarios for the geopolitical environment over the next 5–10 years. Focus on shifts in global alignments, trade, energy security, and technological competition. Avoid any company-specific assumptions.”
“Build three plausible scenarios for the cybersecurity threat landscape facing critical infrastructure in Europe over the next decade. Consider state-sponsored threats, AI-driven attacks, supply-chain vulnerabilities, and regulatory responses.”
“Build three plausible scenarios for sustainability and the energy transition in Europe over the next 10–15 years, considering regulatory pressure, investor expectations, technological maturity, and social acceptance. Focus on the external environment, not on any specific company.”
The goal at this stage is not to test the company’s strategy against these scenarios — that comes later, in Phase 3, once the documentation is on the table. The goal is to enrich the director’s mental map of possible futures, so that when strategic conversations do happen, the range of considered alternatives is already broader and better informed.
Supporting tools in this phase
In practice, this phase can also include:
• Continuous alerts.
• News and media monitoring.
• Regulatory tracking.
• Market and investor monitoring.
• Analyst and sector commentary tracking.
• Real-time research with source support.
• Weak-signal scanning.
• Scenario building based on external sources such as the WEF Global Risks Report or the IEA World Energy Outlook.
The point is to create a background layer of awareness that works quietly throughout the year, not just when board papers arrive.
Phase 2 — Pre-Session Preparation
The second phase begins before the meeting itself, once the agenda and the issues are becoming clearer.
Even when confidential documents cannot be uploaded, AI can still be used as a contextualisation tool. It becomes a way to place internal board discussions against a wider external backdrop: how peer companies are approaching similar issues, what investors are prioritising, what proxy advisers are emphasising, and which governance trends are becoming more visible in the sector.
Tool 4 — The Comparative Lens
“Compare how major [sector] companies in [region] structure innovation governance, digital transformation KPIs, and technology oversight. Identify emerging governance trends, common metrics, and investor expectations.”
That kind of prompt can rapidly produce a comparative framework around a topic that would otherwise require days of fragmented reading.
The same applies to remuneration and governance reviews.
“Compare executive compensation trends, proxy advisor recommendations, and governance sensitivities across major listed [sector] companies in [region]. Highlight areas under increasing shareholder and regulatory scrutiny.”
The value here is not the tool itself. It is the ability to widen the frame before the board discussion begins.
Tool 5 — The Synthesis Workspace
The next tool in this phase is what might be called a synthesis workspace. Sometimes that is ChatGPT. Sometimes it is NotebookLM. Sometimes it is another document-based system or a set of alerting tools that help keep relevant materials current. The point is not to rely on one platform, but to combine tools that help compare, summarise, organise, and revisit information in more usable formats.
This is where a director can build a dedicated topic hub: upload annual reports, governance papers, proxy statements, investor materials, and sector studies, and use the system to identify patterns, recurring themes, and differences across companies.
“Create a comparative analysis of how major European infrastructure companies approach executive compensation, ESG metrics, and long-term incentives. Highlight similarities, divergences, and emerging governance trends.”
That matters because one of the most underestimated constraints in board work today is cognitive bandwidth. Directors are expected to absorb increasingly complex material under severe time pressure, so any tool that helps transform dense information into structured synthesis, comparative tables, or audio summaries can materially improve preparation.
Supporting tools in this phase
In practice, this phase can also include:
• Topic-specific document spaces.
• Audio overviews or podcast-style summaries generated from source material.
• Comparative charts and tables.
• Question-generation prompts.
• Theme clustering across documents.
• Prioritisation of issues before the meeting.
Extra uses in preparation
Some additional ways to use these tools before the session include:
• Building a board question list.
• Identifying what is missing from the pack.
• Extracting the core risks and trade-offs.
• Summarising long reports into a few decision-relevant points.
• Comparing the company against a small set of concrete peers.
• Turning complex documents into audio for revision while travelling.
• Building scenario sets using foresight reports.
• Creating a working expertise space around a strategic topic.
• Generating comparative visuals that make differences visible quickly.
Phase 3 — Decision Reflection and Challenge
The third phase begins once the documentation has arrived and the director is no longer just collecting context, but actively preparing for the boardroom discussion itself.
This is where AI becomes a tool for structured reflection. Most board decisions do not involve obvious right-versus-wrong choices. They involve tensions: resilience versus efficiency, innovation versus control, transparency versus strategic confidentiality, short-term performance versus long-term sustainability. Those tensions are often present, but not always fully surfaced.
Tool 6 — The Trade-Off Mapper
“Map the strategic, operational, reputational, and stakeholder trade-offs embedded in this decision over a 5–10 year horizon. Identify who benefits, who absorbs risk, and what second-order consequences may emerge.”
Used carefully, this does not replace judgment. It helps make the hidden architecture of a decision more visible before the discussion hardens into consensus.
Tool 7 — The Bias Check
This same phase is also where AI can help prepare the debate itself. Many boards do not suffer from a lack of intelligence; they suffer from narrow framing, untested assumptions, or too little constructive challenge. AI can introduce useful friction by generating counterarguments, identifying missing perspectives, and testing whether the issue has been framed too comfortably.
“Argue the opposite side of this strategic position.”
“Assume this decision becomes controversial three years from now. What criticisms are most likely to emerge?”
“Which stakeholder perspective is missing from this discussion?”
“What assumptions are we implicitly taking for granted?”
The point is not whether the tool is “right.” The point is whether it helps widen the range of reflection before the conversation takes place.
Different tools can play different roles here. Real-time source-based systems are especially useful for live exploration, while broader ecosystems become more interesting when combined with documents, workflows, and collaboration environments.
Tool 8 — The Scenario Stress Test
This phase can also include scenario-based challenge. This is where the external scenarios built earlier, in Phase 1, finally meet the company’s strategy. The same tools can be used to test how a recommendation might hold up under different futures — geopolitical, regulatory, technological, environmental — especially if the issue is strategic and long-term. Crucially, this happens at the level of the director’s own thinking, without uploading confidential documents: the scenarios are external, the reflection is internal.
“Given these three external scenarios for the European energy and geopolitical environment over the next decade, what kinds of strategic recommendations would be more or less robust? Identify common vulnerabilities, second-order effects, and the type of decisions that would need to be revisited under each scenario.”
“Assume the external environment shifts materially over the next three years along the lines of these scenarios. Which kinds of strategic commitments tend to age well, and which tend to become fragile?”
That is useful because boards often need not just opinions, but resilience in the face of uncertainty.
Supporting tools in this phase
This is the stage where it becomes useful to:
• Generate counterarguments.
• Test assumptions.
• Surface hidden trade-offs.
• Identify blind spots.
• Build alternative scenarios.
• Prepare questions for the board debate.
• Challenge the logic of the recommendation before the meeting.
• Stress-test strategic choices under different futures.
• Detect missing stakeholder perspectives.
• Examine whether the recommendation is robust or fragile.
Phase 4 — Post-Decision Sensemaking
The final phase begins once the decision is public.
This is where the question shifts from “What did the board decide?” to “How is that decision being read?” Companies do not operate only in markets. They operate in overlapping narrative systems shaped by investors, analysts, regulators, media, political actors, and civil society.
Tool 9 — The Narrative Lens
This is where a mix of tools often works best. A general-purpose assistant can help interpret patterns and build narrative summaries, while other systems can create alerts around media coverage, analyst commentary, stakeholder reactions, and shifts in public discussion.
“Analyse how investors, regulators, analysts, media, and civil society organisations are interpreting this strategic plan. Identify dominant narratives, reputational tensions, and potential areas of misunderstanding or criticism.”
“Compare the company’s stated strategic priorities with external analyst commentary and media interpretation following the earnings release.”
AI can also help examine how narrative shifts interact with market behaviour over time.
“Analyse how market sentiment, analyst commentary, and media narratives evolved following the company’s strategic announcement. Identify correlations between narrative shifts, stakeholder reactions, and market behaviour.”
The point is to understand whether the narrative the company intended to communicate is actually the narrative being received.
Tool 10 — The Stakeholder Feedback Loop
At that stage, the value is no longer just interpretation. It is learning.
Boards can use this final phase to compare how different stakeholders reacted: investors, analysts, media, regulators, employees, civil society, and sometimes customers. That comparison matters because the same decision can be read very differently depending on the audience.
The point is not just to know whether there was a reaction, but to understand the shape of the reaction:
• What did investors emphasise?
• What did analysts question?
• How did media frame the decision?
• Did regulators focus on risk, compliance, or precedent?
• Did civil society raise concerns the board had not fully anticipated?
• Was the intended message received as intended, or was it reframed externally?
In that sense, post-decision sensemaking becomes part of a broader feedback loop.
What matters, ultimately, is not mastering one particular tool. It is building a disciplined way of learning from a family of tools while confidentiality boundaries remain intact.
Perhaps the first real wave of AI in governance will not arrive through fully automated boardrooms or sophisticated governance platforms. It may arrive more quietly, through directors who begin, individually and responsibly, to build their own systems of strategic awareness, comparative preparation, structured challenge, narrative reading, and post-decision feedback.
Governance Thinking Tools at a Glance
| Phase | Tool | Main objective |
| Continuous Strategic Awareness and Foresight | Tool 1 — The Strategic Radar | Maintain continuous situational awareness throughout the year |
| Continuous Strategic Awareness and Foresight | Tool 2 — The Weak Signal Scanner | Identify emerging risks and long-term disruptions across domains |
| Continuous Strategic Awareness and Foresight | Tool 3 — The External Scenario Builder | Build plausible external scenarios (geopolitics, cyber, sustainability, regulation) without using internal information |
| Pre-Session Preparation | Tool 4 — The Comparative Lens | Compare governance practices, KPIs, and peer approaches |
| Pre-Session Preparation | Tool 5 — The Synthesis Workspace | Organise complex information into structured synthesis or audio summaries |
| Decision Reflection and Challenge | Tool 6 — The Trade-Off Mapper | Surface strategic, operational, and stakeholder trade-offs |
| Decision Reflection and Challenge | Tool 7 — The Bias Check | Test assumptions and widen the range of perspectives |
| Decision Reflection and Challenge | Tool 8 — The Scenario Stress Test | Test strategic thinking against the external scenarios developed in Phase 1 |
| Post-Decision Sensemaking | Tool 9 — The Narrative Lens | Analyse external interpretation and reputational tensions |
| Post-Decision Sensemaking | Tool 10 — The Stakeholder Feedback Loop | Compare how different stakeholder groups interpreted the decision |