AI in the Boardroom: 98% of Mid-Market Companies Debating AI-Assisted Decisions
A new Board Value Index survey finds that 98% of mid-market companies have discussed whether to use AI in boardroom decisions. Nearly half are already moving toward implementation, while directors admit boards struggle with rigid processes, information overloa
For a board packed with part-time directors and shrinking margins for decision-making, the question isn’t whether AI can help—it’s who gets to decide how far it goes.
Board Intelligence’s Summer 2026 Board Value Index puts that dilemma in stark numbers: 98% of mid-market companies have discussed whether to involve AI in boardroom decision-making. Of those. 49% of respondents have moved to the implementation stage. meaning they’re actively discussing which parts of the boardroom process could be delegated to AI. Another 34% have discussed AI at the board level without formally proposing specific changes. while 15% have left the question of potential action to company management.
The survey canvassed just over 400 board directors, CEOs and CFOs across the UK, the US, Scandinavia and the Middle East. The companies they represent have annual revenues ranging from £50 million (about $66.8 million) to over £500 million (about $668 million).
What makes the debate more urgent is the mismatch between what boards are expected to do and how directors describe what they actually do. Board Intelligence’s data finds that only 37% of directors or execs describe their boards as “an essential tool for value creation.” And for innovation. directors don’t sound fully convinced: 79% say their boards enable innovation. but only 18% agree they “strongly” enable innovation.
The friction is showing up in process, not just goals. Within the past six months, 86% of respondents say “overly rigid or inconsistent processes” have resulted in poor, delayed or rushed decisions. Forty-one percent say their boards spend too much time discussing past performance.
Against that backdrop. the prospect of AI entering board discussions starts to feel less like a tech experiment and more like a response to everyday constraints. Board Intelligence Senior Director Megan Pantelides points to the kinds of problems AI can realistically address: information overload—processing information more quickly and spotting trends. patterns. risks that might otherwise be overlooked; information asymmetry versus management; inadequate preparation time. given that most board members work part-time across multiple boards; and cognitive bias.
Pantelides argues these issues have “a direct impact” on the quality and robustness of board decisions, and that AI offers scalable methods for dealing with them.
In her view. AI’s role—at least for now—is less about “taking over” and more about doing the work that tends to get squeezed first. It performs the “analytical heavy lifting” that can inform board decisions. That includes “synthesising large volumes of information. ” surfacing patterns across data. and flagging risks and assumptions that might otherwise be buried in a 200-page board pack.
There’s still a curious gap between the enthusiasm and the evidence. Pantelides says there aren’t many specific examples of named companies actively using AI—or openly admitting to using AI—in board decisions. Still. she does cite Abu Dhabi-based sovereign wealth fund Mubadala as one organization with an AI member on its investment committee.
Lloyds Banking Group is one of the clearer public examples. In April. it began using Board Intelligence’s automated tools to help executives prepare for meetings. primarily through analysis and summarisation of documentation. Pantelides says Lloyds could potentially give AI a greater role later—using it to identify and reduce bias in decision-making and provide informational support across a range of areas. including cybersecurity. M&A activity. sustainability. and market analysis.
The deeper tension sits in what boards are willing to hand off. Pantelides suggests AI agents may take time to reach the point of making decisions rather than informing them. She links the caution to something boards are already signaling: the high proportion of boards discussing which decisions should remain human-led.
Her line is blunt—boards are grappling with where the boundary should lie between AI-assisted analysis and human judgment. Legal accountability. she says. will always likely remain with human board members. and any directors trying to keep litigation risk down would be wise to double check and authorize recommended actions themselves.
That boundary matters because the types of decisions boards make vary wildly in complexity and consequence. Pantelides places policy-change sign-offs at one end of the spectrum and decisions involving M&A. culture. leadership. strategy. and crisis response at the other. In her words, boards are not prepared to delegate the latter—and nor should they.
Taken together. the survey reads like a moment of pivot: 98% of mid-market companies are debating AI-assisted board decisions. almost half are already in implementation conversations. and the stated motivation is tightly tied to what directors say is going wrong today—rigid processes. rushed outcomes. slow digestion of information. and the biases that slip into judgment.
The argument now isn’t whether AI can help boards see more clearly. It’s whether that clarity can stay advisory—before the day comes when the board room asks AI not just to analyze, but to decide.
AI boardroom decisions Board Value Index Summer 2026 Board Intelligence Megan Pantelides mid-market companies board governance information overload cognitive bias Lloyds Banking Group Mubadala cybersecurity M&A sustainability