Foreword

AI represents an age-defining opportunity for society and business. And right now, most companies are moving at the pace of organizational change, not the pace of the technology itself. The spotlight is on organizations to close the gap and rapidly harness the full potential of new technologies in a way that both enhances their performance and positively impacts their stakeholders.

Balancing the opportunities and risks of AI is therefore becoming central to boards’ governance and stewardship responsibilities. Their oversight can ensure AI adoption drives performance and resilience while aligning with organizations’ ethical standards and long-term vision for a trustworthy future.

The stakes are high. Poorly implemented AI can have far-reaching consequences, eroding trust and damaging reputations. Conversely, TrustworthyAI^TM can transform organizations and communities for the better.

For many organizations, this transformation is still in its early stages—but overall, awareness of AI and the importance of its oversight has grown. Findings from this second edition of Governance of AI: A critical imperative for today’s boards show a growing impetus for action. Boards recognize the need to accelerate education and adoption and are making strides toward embedding AI in their agendas and in their conversations with a broader set of C-suite executives.

Despite this progress, the results also indicate that the focus on AI governance should accelerate. Technology evolution was a challenge when it was measured in years, but the advancement of AI is happening much faster, including the shift toward agentic and physical AI. Boards that recognize the opportunities and risks AI brings, and their responsibility in governing them, can help position their organizations for success in this new and exciting future.

To help board members embrace this responsibility and guide their organizations forward, this report proposes key questions they should ask of themselves and their management teams. As Chairs and boards, we may not have all the answers, but our strength lies in asking the right questions.

Anna Marks
Chair
Deloitte Global Board of Directors

Lara Abrash
Chair
Deloitte US

Defining the opportunity: Key themes from the survey

In the context of this report, the term “AI” is used broadly to encompass all forms of artificial intelligence, including but not limited to Generative AI. This includes machine learning, natural language processing, computer vision, and other AI technologies.

Slowly but surely: AI increasingly on the board agenda

More boards are putting AI on the agenda. In this latest survey, almost one-third of respondents (31%) said AI is not on the board agenda, which represents a 14 percentage point decrease from the previous edition. While fewer than one in five boards (17%) address it at every meeting and 19 percent take it up only once per year, this is still progress. In a rapidly evolving digital landscape, proactive board engagement on AI will likely be critical. As AI becomes an integral part of business strategy, boards that are slow to prioritize AI on the agenda risk falling behind or missing out on opportunities for innovation, efficiency, and market leadership.

Board considerations for the boardroom agenda: Overseeing a company’s AI adoption and its related opportunities and risks requires a board to first understand the company’s current AI maturity. Questions for the board to consider in this area, especially if they are looking to add it to the board agenda, might include:

Pace of AI adoption: Boards see the need for acceleration

At a time where technological advancements, commercial applications, and regulatory activities are rapidly evolving, it is crucial for organizations to adopt at pace. The latest survey findings underscore the importance of prioritizing AI adoption to harness AI’s transformative potential for operational efficiency and strategic advantage.

How do board members and top executives assess their own organizations’ pace of AI adoption? Just over half (53%) of respondents feel the need to accelerate efforts while only one-quarter (25%) are “satisfied” or “very satisfied.”

What about progress to date? Only 5 percent of respondents report that AI is incorporated into their business and operating plans going forward. Nearly 40 percent are experimenting with the technology (38%) or focusing in specific areas (35%). In the previous survey, about one-third (32%) said AI was not incorporated into their operating plans at all, but this time only about one-quarter (22%) said the same.

In terms of preparedness, almost one-third of respondents (31%) say their organizations are not ready to deploy AI. In the previous survey, that figure was 41 percent, which marks a notable increase in preparedness levels. About one-third (32%) are “somewhat ready” and only 5 percent are “very ready,” both of which are slight increases from the previous survey. Similarly, Deloitte’s State of Generative AI in the Enterprise Q4 2024 report report finds that organizations have improved their preparedness in the critical areas of technology infrastructure and strategy, but preparedness has seemingly not improved in the critical area of risk and governance.

Board considerations for the pace of adoption: Boards can start by thinking about their role in governing strategy execution and risk management—overseeing and challenging management to identify when and how the strategy may need to be adapted in response to risks and opportunities. Questions for the board to consider include:

Opening the communication channels

While the data shows boards are willing to take the initiative on AI, they cannot take a siloed approach in doing so. By interacting with executive leaders in key parts of the organization, boards can help connect direction and strategy with actions and outcomes. The latest pulse survey finds that AI-related interactions between the board and the C-suite are strong when it comes to the Chief Executive Officer (CEO) and executives responsible for technology, and growing more frequent with other CXOs.

When asked which management team members their boards were talking to about AI, an overwhelming majority of almost three-quarters (72%) mentioned the Chief Information Officer (CIO) and Chief Technology Officer (CTO), and over half mentioned the CEO. Those roughly track with the previous survey’s replies.

The new survey also shows continued interaction with other top executives, which could indicate a recognition that AI strategy requires a holistic, integrated viewpoint. For example, 27 percent of boards engage on AI with their Chief Financial Officers (CFOs) and 12 percent engage with their Chief Information Security Officers (CISOs) and Chief Risk Officers (CROs) respectively. Only 8 percent of respondents said their boards do not engage with management on AI at all, down by 5 percentage points from the last survey.

As elsewhere, much of the story is still unfolding, but more top organizational leaders are having these conversations. By fostering these conversations, boards can move AI initiatives from theoretical discussions to practical applications. Failure to open these communication channels could result in fragmented strategies or an inability to fully leverage AI’s potential within the organization.

Board considerations for opening communication channels: In governing AI, the board can work with management and oversee the governance framework, oversight roles, and leadership and management structure. Questions for the board to consider include:

Board members show curiosity

In light of the finding that two-thirds of board members and executives in the survey reported limited to no AI knowledge and experience, there appears to be a case for prioritizing more learning at the leadership level. Roughly half of the leaders surveyed said their organizations are providing foundational AI education to their boards—which, even though it represents an 8 percentage point increase from the previous survey, still means the other half are not.

At the same time, many individual board members are taking the initiative: 59 percent of individual board members report they’ve sought to enhance their AI knowledge on their own. That aligns with the low percentage (25%) of board members who are satisfied with AI’s place on the agenda. It might indicate welcome initiative, but it may suggest that AI training and education are not being coordinated in a consistent way. Without a comprehensive approach to AI education, boards may struggle to fully grasp AI’s strategic importance. Ensuring all board members are well-versed in AI will be crucial for driving informed board discussion and decision-making.

Board considerations for AI fluency: Boards will need to stay informed on AI developments relevant to their industry and the company specifically. Questions for the board to consider include:

Navigating the path forward

This pulse survey shows that more boards are placing AI firmly on the agenda, are keen to see the pace of adoption in their organizations accelerate, and are spending more time educating themselves on the topic. Still, the findings indicate that work remains to be done in most organizations for them to effectively oversee this technology. As in other areas, boards that consider input from experts inside and outside their walls to help inform critical decisions can complement that expertise with knowledge and direction of their own.

The board’s responsibility is to sustain the organization’s long-term value and strengthen its resilience. Oversight of this technological revolution is a vital part of that responsibility. As technology continues to evolve, boards, together with the C-suite, have the chance to take a more comprehensive look at AI opportunities and risks to build more innovative and resilient organizations.

The considerations in this report have been derived from the AI Governance Roadmap developed by Deloitte US. The Roadmap offers key insights and questions to help oversee AI risks and opportunities, promote ethical use, and navigate the evolving landscape. Other considerations for boards on this topic include: