A BCG survey reveals that 61% of CEOs believe boards are hastily pursuing AI transformation due to the distortions caused by hype in boardroom decision-making.

      A BCG survey involving 625 CEOs and board members revealed that 61% of chief executives feel their boards are moving too quickly on AI transformation. While 75% of board members consider their AI knowledge to be adequate, almost 40% of CEOs disagree, with over half asserting that the hype surrounding AI is affecting board decisions.

      The survey, titled "Split Decisions," included responses from 351 CEOs and 274 board members from companies with at least $100 million in annual revenue, highlighting a clear divide: both boards and CEOs agree on the significance of AI, but they differ on the pace of its implementation, the depth of the boards' understanding of it, and how much a CEO's role involves generating returns from AI.

      These findings come at a time when fear of missing out (FOMO) on AI has become a significant influence on corporate strategies. More than half of the surveyed CEOs believe that the buzz around artificial intelligence is clouding their boards’ judgment, and nearly 40% think their boards are uninformed about AI's impact on growth strategies. Additionally, one-third of the CEOs feel that their boards overestimate the extent to which AI can replace human roles.

      The survey highlights a notable disconnect between how board members assess their own knowledge of AI and how their CEOs perceive it. Three-quarters of board members rated their understanding as comparable to or better than their peers, while CEOs were significantly less impressed. This suggests that many boards may be making important decisions regarding AI strategy based on what their CEOs consider insufficient knowledge.

      Julie Bedard, a managing director and partner at BCG, noted that this knowledge gap can be bridged if CEOs take it upon themselves to educate their boards. Instead of relying on technology officers or external consultants for AI briefings, she believes CEOs should lead training sessions that clarify what current AI tools can and cannot do. It is important for these sessions to differentiate between functions where AI replaces human effort and those where it complements it.

      This distinction is crucial; boards that view AI solely as a replacement for human labor may advocate for quicker and broader deployment than the technology can feasibly handle. Conversely, boards that see AI as a complement to human work are more likely to endorse investments with achievable outcomes. The survey suggests that too many boards fall into the former category, with the negative consequences of FOMO-driven investments in AI becoming increasingly evident.

      The survey also revealed a discrepancy in how CEOs and board members perceive accountability for AI outcomes. CEOs estimated that 35% of their performance reviews now hinge on achieving returns from AI, while board members estimated this figure at 27%. This eight-percentage-point gap implies that CEOs may feel a stronger pressure to deliver AI results than boards recognize.

      This has implications for behavior; a CEO who believes that over a third of their evaluation depends on AI results will be incentivized to prioritize AI initiatives, even if those initiatives are hasty or poorly defined. Conversely, a board that perceives this figure as lower may be confused about why its CEO is hesitant to act more swiftly or might underestimate the operational risks of hastening implementation to meet perceived expectations.

      Judith Wallenstein, BCG’s managing director and senior partner leading its global CEO Advisory practice, emphasized that CEOs must bring their boards along on the same learning journey they have undertaken, albeit in a more condensed and focused manner aimed at fostering true understanding rather than superficial awareness. The practical and operational realities of deploying AI are often far more complex than the presentations typically seen in boardrooms prior to investment decisions.

      It's important to note what the survey does not address. It does not assess whether the CEOs who feel their boards are moving too fast are justified in their caution or if certain boards are correct in urging faster action. In some industries, rapid AI adoption may be the optimal strategy, and the resistance from CEOs could stem from organizational inertia rather than sound reasoning. Thus, the survey reflects a perception gap rather than a determination of who is right.

      The study also does not provide a breakdown of results by industry, geography, or company size beyond the $100 million revenue threshold, limiting the insights that can be drawn for specific sectors. For example, a board advocating for AI transformation in a financial services company faces different risks compared to a board in a manufacturing firm, yet the survey treats both scenarios equally.

      What the research does confirm is that senior leaders at large companies are not on the same page regarding this crucial technological investment era. Approximately 80% of both CEOs and board members agreed that potential board candidates should demonstrate a measurable understanding of how AI can impact their industry, indicating an awareness of the knowledge gap despite differing opinions on its severity.

      A deeper question raised by the survey is whether traditional board governance is suitable for making AI-related decisions. Boards typically convene a few times a year, rely on management presentations for insights, and consist of members whose expertise may focus on finance, regulation, or specific operational areas rather than technology. This structure worked well when technology evolved at a pace that allowed