AI fluency at the board level is the practical ability of directors to evaluate what they are being told about artificial intelligence, ask the questions that surface real risk, and make governance decisions that hold up under scrutiny. It is not technical expertise. Boards are not being asked to build AI systems or understand the mathematics behind them. They are being asked to govern the strategy they have already funded, and to do that well, they need more than a briefing. They need to be AI fluent.
…
Most boards are governing AI they do not yet fully understand
Most boards today have AI somewhere on their agenda. It shows up in the strategy deck, in the capital allocation discussion, in the risk register. What tends not to show up, at least not yet, is a clear sense of what directors themselves need to understand about AI to govern it credibly.
The range of AI experience in a typical boardroom is wide. While some directors have spent years deploying AI solutions at scale, others opened their first AI account just a few months ago. Both are expected to evaluate the same strategy, challenge the same management presentations, and understand the same risks. The governance expectations placed on both are identical, which is precisely what makes the fluency gap so difficult to close.
This is the gap organizations need to tackle now. Technology is already evolving at a breakneck pace, and the boards that wait will find the distance harder to close. Before getting into how to develop AI fluency, it helps to be clear on what it actually is and what it is not.
Most of the progress boards have made on AI so far has been at the level of literacy. Fluency is a different destination, and the distinction between the two matters more than it might initially appear.
AI literacy vs AI fluency
| AI Literacy | AI Fluency | |
|---|---|---|
| Focus | Concepts, vocabulary, and the risk landscape at a foundational level | Accountability, oversight, and the ability to evaluate specific governance questions |
| What a director can do | Follow AI conversations and engage without appearing uninformed | Challenge management, ask accountability questions, and identify gaps in what is being reported |
| In the boardroom | Vocabulary and general awareness of AI risks | Judgment to govern AI decisions rather than ratify them |
| Where the risk lies | Easily mistaken for competency; leaves judgment gaps that surface when things go wrong | Requires ongoing development, not a one-time achievement |
A board training session that produces literacy can give directors the impression that they have what they need, when what it has actually built is confidence without the underlying judgment. When something goes wrong, directors may find they were working with an incomplete picture, and that it shaped decisions they are now accountable for.
Vocabulary is neither governance, nor oversight
Understanding AI vocabulary and being able to place it in the right conversations is a positive sign that the board has completed meaningful introductory training. But if directors stay silent when the time comes to ask questions about oversight and accountability, it signals that a critical part of the picture was never addressed.
Questions about which data management processes run through specific tools, the liability exposure if an AI-assisted decision is challenged, and whether the organization’s D&O coverage accounts for AI in the decision chain must come from the boardroom. A board that cannot ask them is ratifying rather than governing, and that opens the door to a different set of challenges.
The risk of AI washing grows here too. A director who cannot evaluate what the organization is publicly claiming about their AI abilities cannot catch inconsistencies before they become enforcement problems. The SEC, DOJ, and FTC have all acted against companies that overstated what their AI does, and preventing it requires someone in the room who understands the gap between what is being said and what is actually happening.
“Shadow AI” is already invading the boardroom
More and more boards acknowledge that AI has to be part of the governance agenda and are beginning to act on it. There is an external issue, though: a significant number of directors are already using AI through personal accounts, with no policies in place, no secure material sharing, and no guardrails. The governance gap in this case goes beyond what directors oversee in the business. It extends into how they use AI in their own work.
Effective AI training for boards helps mitigate these risks by ensuring the boardroom understands the ethics and principles of responsible AI use. Having a clear AI code of conduct also plays a role here, giving directors explicit guidance on what sanctioned use looks like before informal habits take hold. A single training session cannot replace recurring, ongoing AI education. The boardroom is expected to hold accountability on AI decisions, and that accountability has to be maintained as the technology, the regulatory environment, and the organization’s own AI use all continue to evolve.
Why AI fluency matters
To understand why AI fluency matters as much as AI literacy, the boardroom needs to think beyond investor pressure.
The regulatory challenge
The conversation is often framed around investor pressure and regulatory compliance. The Anteris Advisors 2026 Proxy Season Review, published by the Harvard Law School Forum on Corporate Governance, found that AI has emerged as a fast-growing topic in shareholder proposals this year, drawing growing institutional attention. That is a meaningful signal about where the accountability conversation is heading, but it is not the whole picture.
In Europe, AI readiness might become a formal governance requirement. The EU AI Act and evolving national implementations are moving toward a landscape where the competence to govern AI could factor into the fitness and propriety standards applied to board members, a standard already established in financial services. Nothing is codified yet, but the direction is clear enough that a board building fluency today is in a stronger position than one waiting for formal requirements to arrive.
Can AI make your board meetings more efficient?
Discover three practical and secure ways to streamline board meeting preparation and follow-up, improve decision-making, and protect sensitive data with AI-driven tools.
The leadership challenge
Beyond regulation, the more immediate consequence of the fluency gap is what it does to the board’s ability to actually lead. A board that cannot evaluate AI proposals confidently tends toward one of two responses: approving whatever management brings without meaningful challenge. Or it creates friction and delay out of unexamined caution. Neither serves the organization. In both cases, decisions about AI strategy, investment, and risk management are being made without the board contributing meaningfully to them.
That matters more than it might appear because the board sets the tone for how seriously AI governance is taken across the organization. When directors demonstrate real fluency, it signals to management that AI decisions will be examined, that they have taken responsibility for the outcomes, and that the organization’s approach to AI can withstand scrutiny. That signal travels as it shapes how risk is assessed at every level below the board, how quickly new AI initiatives move, and how well the organization builds a track record it can stand behind with investors, regulators, and customers alike.
What board-level AI fluency looks like
To reach that point, the boardroom needs to understand what is expected of them. It goes well beyond knowing their AI terminology and its implications.
The ability to ask the right questions
Not “how does the model work?” but “who owns this output if it is wrong?” and “what happens if this decision gets challenged?” A fluent director redirects technical explanations toward accountability, and knows which answers are good enough and which ones are not.
Awareness of what the organization is saying about AI publicly
Filings, marketing materials, customer communications, and sales conversations can all create inconsistency that becomes legal exposure. A director with no working knowledge of the company’s AI claims cannot contribute to preventing the gaps that regulators are now actively looking for.
A sense of the risks and exposures of different AI usages
It does not have to be a comprehensive legal map, but enough to know that an AI tool touching hiring decisions carries discrimination risk, that one involved in financial projections carries disclosure obligations, and that one generating legal content carries hallucination risk that requires human review. This is governance fluency, not technical fluency, and it is entirely learnable.
The personal discipline to use AI appropriately in their own role
A director who processes board materials through an unsanctioned personal tool undermines the governance structure the board is trying to build. Fluency starts with knowing what responsible use looks like for yourself, before expecting it from the organization you oversee.
Ongoing education as AI continues to develop
Fluency is not achieved through a single training session. It requires a continuing commitment to learning, practice, and honest assessment as AI capabilities, regulations, and organizational use cases evolve. Directors should revisit the fundamentals regularly, but they also need opportunities to apply them to the decisions in front of the board: evaluating an AI-related investment, challenging a risk assessment, reviewing an incident, or testing whether management can explain accountability clearly. The objective is to keep judgment current, so directors can recognize when a familiar risk has changed, when a new use case needs deeper scrutiny, and when management’s assurances are not supported by sufficient evidence.
Fluency fit for the future
Most governance conversations about AI still frame the problem as catching up, about getting boards to the point where they can follow along. The problem is that bar is now too low. The technology is embedded deeply enough, and the decisions that flow from it are consequential enough, that following along cannot be the end goal.
The organizations getting this right are building a boardroom that knows when to push back, what to ask for, and what it means when answers do not hold together. To sum it up: they are building governance skills. Just like other governance skills, AI fluency also builds through practice, honest assessments and commitment to close the gaps before they become problems. Boards that understand how AI is embedded across the organization are better placed to build governance that keeps pace with the technology.
Frequently Asked Questions
Board AI fluency is the ability of directors to evaluate AI-related information, ask meaningful questions, identify accountability gaps, and make sound governance decisions. AI literacy provides a foundation in AI concepts and risks, while AI fluency helps directors challenge management and oversee how AI is used across the organization.
Boards are already making decisions about AI investment, strategy, compliance, and risk. Without sufficient fluency, directors may approve management proposals without proper challenge or delay decisions because they cannot assess the issues clearly. AI fluency helps boards examine questions about data, liability, bias, public claims, cybersecurity, human oversight, and the impact of AI-assisted decisions
Boards can build AI fluency through recurring education, practical scenarios, and focused discussions linked to the organization’s actual use of AI. Directors should review how AI affects strategy and risk, clarify accountability, understand approved and prohibited uses, and revisit their knowledge as the technology, regulations, and business applications change. One training session can introduce AI literacy, but sustained learning is needed to develop lasting governance judgment
