Although anecdotal, this survey of directors shows both the limitations and the potential of AI to assist boards of directors. Using AI for information gathering and analysis is not widespread among boards. However, there are indications that this trend is catching on as part of the overall inroads AI is making into organizations, from the mailroom to the boardroom. 

There is a massive potential for AI to change the fundamental way an organization is run. Nearly 80% of companies in a recent McKinsey survey reported using AI in areas such as workflows, business processes, and data generation and analysis. Already, boards are devoting more time and attention to discussing the use of AI within their organizations, including more than 62% of directors who reported participating in full-board discussions about AI policies for their companies.

Now, the question becomes how board members can use this technology—efficiently and ethically. As a long-time board member, and a former chair and CEO of a $12 billion health care company, I am optimistic about AI as a contributor to problem-solving and decision-making. Such capabilities increase the likelihood that boards and management teams, alike, will adopt this technology—not as a replacement for human judgment, but as a tool that enables more informed decision-making.

Augmenting board business

AI has the potential to help board members become more knowledgeable and better equipped to engage in discussions with management about challenges, opportunities, strategy, and operational issues. Its biggest contribution for directors could be in supplementing the “board package” delivered throughout the year by the company to summarize recent performance, new developments, competitive challenges, and potential acquisitions. Board members are expected to read and digest this information before the next meeting so they can ask questions of management and engage in discussion. This information, however, is fed to the board by the company, which by necessity involves a degree of filtering to decide the amount of detail required and the depth of the discussion. 

Open, honest, and transparent information—thorough enough for meaningful discussion without inundating directors with details—is a cornerstone to maintaining a trusting relationship between the board and management. As a board member told me when I first became CEO of Baxter International, “Never surprise me, Harry. I don’t want to be driving my car and hear news about Baxter on the radio that I wasn’t aware of.”

However, even with the best of intentions, management cannot be expected to educate board members on every economic, geopolitical, or operational issue impacting the company. Additionally, it may be tempting for management to focus more on what’s going well, rather than presenting to the board some of the significant issues the company is facing. In an increasingly complex world, board members need to take more responsibility to educate themselves and to deepen their understanding of the many sides of an issue—cultivating a balanced perspective that is crucial to values-based leadership. In these pursuits, next-generation AI tools can be a tremendous help. 

Consider the example of a company that is weighing the impact of U.S. tariffs on goods imported from China and the potential to shift production to locations elsewhere in Southeast Asia. With trade talks evolving rapidly, tariffs are being proposed, rescinded, and amended quickly across the region. Quick and concise AI queries can help board members aggregate information about the current state of tariffs and trade talks. With a few keystrokes, board members can also gain insights into how other industries are being impacted and what production decisions they are making. 

The human part is tricky

A tool, however, is only as good as the user. As my colleague at Northwestern University’s Kellogg School of Management Mohan Sawhney once observed, the technology aspect of using AI is comparatively easy; rather “it’s the human part that you really need to pay attention to, because if you don’t think about humans, machines won’t think about you.”

There can often be an enormous educational requirement to help board members become more comfortable using AI. The average age of board members for an S&P 500 company is 63, with many serving until their 70s—meaning they’re hardly digital natives. In addition, as anyone who has used generative AI knows, the quality of the answer received depends very much on the quality of the question being posed. This will require far more than a one-hour presentation by a consultant to the board. What’s needed is training that will help directors put this tool into practice, with considerations around ethical use such as not disclosing confidential company information while posing a query to AI. 

As the business landscape continues to evolve, companies and their boards need to keep pace with the change. Board members can better serve shareholders and carry out their fiduciary duties by learning how to use all the tools—including AI—at their disposal. 

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.