The board question comes before the meeting agenda closes. It comes in the Q’A, after the CFO’s slides, when someone at the end of the table looks up and says: “So what is our AI strategy?”
The pressure in that room is real. PE-backed companies now report that 30 to 40 percent of investor and board discussion centers on AI. Some sponsors have announced fund-level AI deployment commitments in the billions. The question is no longer whether AI belongs on the board agenda. It is whether you can answer it credibly when it lands.
I have sat in those rooms from both sides, as an operator and as someone who has been inside PE-backed companies doing the actual AI implementation. Here are the five questions I hear most consistently, why they get asked, and what a credible answer actually looks like.
1. What is our AI strategy?
Why boards ask it: They are not asking for a philosophy. They are asking because every IC memo they have reviewed recently includes an AI section, and they want to know if this company has one too. Investors and board members are now fielding questions from their own LPs and stakeholders about how portfolio companies are responding to AI. “We are exploring it” does not answer the question.
What a credible answer looks like: A real AI strategy names three things. First, the specific workflows or domains you are redesigning, not the tools you are buying. Second, the measurable outcomes you expect, whether that is a reduction in financial close time, faster customer onboarding, or lower cost per service ticket. Third, the person who owns it. If your AI strategy lives in a slide deck with no named owner and no metrics, it is not a strategy. The companies I see making real progress have one workflow per quarter on a clear implementation plan, not a portfolio of experiments.
2. How do we compare to peers in our sector?
Why boards ask it: Competitive anxiety is legitimate. Boards have access to peer benchmarks, industry reports, and investor portfolio data that their management teams sometimes do not. When a board member sits on three other portfolio company boards and sees one of them automating its revenue cycle while yours still takes four people five days to close the books, they notice. The peer question is often a signal that they already know the answer and want to see if you do too.
What a credible answer looks like: You need actual data here, not a vague sense of where you stand. This means knowing your key operational metrics, financial close time, onboarding cycle, headcount-to-revenue ratio, and how they compare to industry benchmarks. It also means being honest. If a competitor has moved faster, say so, and name what you are doing about it. Boards trust executives who diagnose accurately. They lose confidence in executives who rationalize.
3. Where should we invest first?
Why boards ask it: Because everyone is asking for budget and the board needs a framework to evaluate competing priorities. They have heard pitches for AI tools, data platform upgrades, and training programs, and they need someone to tell them which one is the actual bottleneck. This question is often a test of whether leadership understands the dependency chain.
What a credible answer looks like: The honest answer for most mid-market companies is data infrastructure and governance first, then use cases. You cannot build reliable AI on unreliable data. A mid-market industrials company we worked with had already purchased AI tooling before we got there. The tools were fine. The problem was that their revenue data lived in three systems with no single owner and no reconciliation process. Every AI output required manual validation. The tools were not the bottleneck. The data foundation was. Invest in the foundation that makes the use case trustworthy, then build the use case.
4. What are the risks we are not managing?
Why boards ask it: Governance and risk management are core board functions, and AI introduces a new class of risk that most mid-market companies have not formally assessed. From a fiduciary standpoint, a board that has not asked about AI risk is not doing its job. They are also watching what happens at other companies: AI-generated financial models that look finished but contain errors, compliance exposures from automated customer communications, and the liability questions that come with AI-assisted hiring or pricing decisions.
What a credible answer looks like: You need an honest inventory. The three risk categories I see most consistently underweighted in mid-market companies are output verification (who checks what the AI produces before it informs a decision), data exposure (what customer or employee data is being sent to third-party AI systems), and agent ownership (if AI is running automated workflows, who is accountable when something goes wrong). A company that cannot name its AI risks is not managing them. A company that has assessed them and built a governance process around them has something durable to show the board.
5. How do we measure progress?
Why boards ask it: Boards that have funded AI initiatives are now asking for the return. The license cost shows up on the income statement. The value should show up somewhere too. “We have increased adoption” is not a metric. Neither is “the team is excited about it.”
What a credible answer looks like: Tie AI metrics to business outcomes that already appear in your reporting. If you are automating accounts receivable, measure days sales outstanding. If you are using AI in customer support, measure cost per ticket and resolution time. If you are deploying it in recruiting, measure time-to-hire and offer-acceptance rate. The companies that sustain board confidence on AI are the ones that can point to a metric that moved and explain why AI moved it. Everything else is theater.
CTA
Most executives I work with are not lacking ambition on AI. They are lacking a structured way to answer these questions before the board asks them. Blueprint is a tool I built for exactly this: a 15-minute AI readiness assessment that gives you a score across your data infrastructure, talent, governance, and use case pipeline, with a prioritized roadmap for where to invest next.
If your next board meeting is in the next 90 days, run it now: blueprint.blueorange.digital