While artificial intelligence is rapidly reshaping the way boards access information, assess risk, and guide strategy, its role in the boardroom is actually still taking shape. For Jeanne Beliveau-Dun, an independent director at Columbus McKinnon, Edison International and Crudl, this shift is less about adopting flashy new tools and more about embedding AI into core areas of governance such as oversight, risk management and long-term value creation. She sees first-hand how boards are beginning to carefully, and often unevenly, integrate AI into their workflows across sectors with vastly different regulatory and operational pressures.
Beliveau-Dunn suggests that what distinguishes effective boards today is not how aggressively they deploy AI, but how thoughtfully they address its impact. Directors are weighing the benefits of faster insights and deeper analysis against practical concerns about cybersecurity, data integrity, and legal risk. At the same time, there is increasing pressure for organizations to keep up in an increasingly AI-driven competitive environment.
What specific AI applications are boards actually using today? Which do you think are still overlooked or undervalued by most directors?
Traditional board portals like Diligent all have AI versions that help summarize data, identify trends, and assist with workflows. Additionally, many large boards have access to company-specific LLMs, allowing them to safely research company topics and company data in their LLMS. While some boards allow the use of meeting summary tools (often from private companies and tech companies) for board meetings on Zoom, Microsoft, or third-party apps, many boards are wary of using them legitimately for legal reasons. It is important to constantly measure the risks and rewards of the usefulness of tools and the legal risks of storing more information and data online that have not been previously recorded in detail. It really depends on the size and department of the board. Regulated industry sectors have far greater concerns about these issues and are far more cautious than, for example, the technology sector.
How is AI changing the way boards approach risk oversight?
The risk of industry disruption is greater than ever if companies do not effectively leverage technologies such as AI to remain competitive. On the other hand, AI could pose new threats through adversaries expanding their cyberattacks. This allows cybersecurity to avoid attack vectors to businesses, quadrupling the likelihood of a cyber intrusion before it can be contained. The good news is that new AI tools are becoming available in cyber that are much better than in years past. These tools can automate threat detection, perform continuous monitoring, and even automate responses. Similarly, Chat GP Codex/Enterprise can now perform IT code inspection to find exploitable software holes, recommend patches, and automate the process. Anthropic/Claude Code Security has been a leader in this space by discovering coding flaws that identify holes in systems, such as SQL injections, authentication flaws, and the risk of data leaks in existing software and systems that are considered secure. As a result, businesses have had to take urgent steps to prepare for an onslaught of new fixes from vendors who have identified these issues. Cyber is the biggest risk to businesses. AI-based threats to cybersecurity should be near the top of every board meeting agenda. Additionally, using AI outside of locked-down systems poses its own risks, such as loss of privileges due to data breaches or legal disputes.
What are some examples of boards that are using AI effectively, and others that have AI fatigue, or that still struggle to differentiate between hype and real business impact?
The more important question is how companies are leveraging AI. Because the board's job is very limited in manufacturing, and it's more about monitoring, so AI is used to summarize trends, track data, and compare data. This will definitely improve the acquisition of third party information/research on non-confidential topics to educate the board and allow them to become more informed on the topic.
That said, the most important thing for boards and executives is to understand how AI can transform their industry's operational capabilities and products/services, and how board members should manage the use of AI in those areas. We are seeing many companies move from experimentation to now prioritizing key areas of business improvement/product/service superiority. In some cases, key internal strategy leaders will need to lead this effort. You can also find the highest value applications/use cases that should be prioritized and funded with the right set of metrics for governance, with the help of knowledgeable external consultants with experience in that industry.
AI can quickly synthesize large amounts of data. How should boards leverage their capabilities to make better strategic decisions without becoming overwhelmed or creating a false sense of confidence?
It's really about driving specific reports to the company that use AI to do all kinds of analysis. Analysis topics are very diverse, but some that come to mind include estimating climate risk for business, financial risk for management, analyzing cyber intrusion risk, industry trends over the years regarding the best performing products and categories, streamlining coding and software development in IT, and how to gain back-office benefits through AI. Benchmarking all kinds of KPIs using the right industry data can be very helpful for companies and boards.
Are boards using AI to predict problems before they occur to executives, or are they still mostly reactive? What does proactive governance with AI actually look like?
AI is now being implemented in most audit functions, finance functions, and should be implemented in at least most IT departments, along with marketing and legal departments. Boards should encourage their operational teams to leverage AI-powered data-driven decision-making with the board. We're in the early stages of all of these efforts, but we need to insert this into all major features. However, it is still early days, and there are many vendors on the market that can help identify industry-specific risks and provide proactive data. Datamaran is an example of risk assessment/governance.
How can boards use AI to reduce complexity and speed up investigations in areas like compensation, regulatory compliance, and financial reporting without losing the human judgment that is critical to governance?
First, this is not about automating the director’s job, but rather facilitating informed decision-making through better-informed data available through AI. Most compensation advisory firms working with boards of directors have already begun leveraging AI in their operations to help boards manage compensation-related decisions and benchmark appropriately. This will help streamline data on benchmarking, more quickly analyze how companies compare to each other, and enable better governance in key areas of compensation and development strategy. They can also make summary recommendations based on benchmarks and governance standards, but every board should engage a remuneration advisor who aggregates this third-party data using AI tools. The main reason for this is that decisions are not just about data, but about the specific strategies to deploy for specific companies based on where they are in the industry performance category and how their risk profile has evolved. These consultants should provide AI-driven analysis to recommend several optimal paths/alternatives based on the goals and philosophy of the company and management. AI is already having a huge impact on worker compensation, and I think most companies are leveraging it. A specific example of a company in this category is Compa.AI, which uses AI to compare real-time information on job postings and old statistical data. Some services, like Salary.com and Payscale, use AI for different audiences.
If a board receives an AI-generated report or analysis, what questions should the board ask to validate the results?
They need to ask questions about the source of the data. You should fact-check a possible hallucination problem by running the prompt or query multiple times to see if you get the same answer. We also conduct independent research on non-confidential and open topics such as competition and market trends that allow directors to continue to add value to their boards. IT helps directors see and understand data and better understand the various decision choices they may face. AI is very good at answering “what if” questions and creating different models from questions to outcomes. It's all about the quality of the data and the quality of the analysis, so ask what prompts are used to arrive at the analysis.
What is your advice for boards that want to implement AI tools but lack the technical expertise? How can directors build confidence and competency around AI-driven insights?
First, make sure you have core technical people on your board who have run businesses who can lead this. And everyone should use AI tools in their daily lives to understand their capabilities and impact, as well as what they can't do. The capabilities of this technology are changing every day and continue to improve exponentially. Some level of literacy is required across the board, and professionals who are very knowledgeable about AI and cyber and have extensive business experience should also be considered.
Some boards feel they need to bring in AI experts, but there are concerns that technology experts won't be able to contribute meaningfully to other important governance issues. And, of course, the number of seats on the board is limited. What do you think about that?
You're right that unless you're an AI company, you don't need or want pure AI engineering or research people on your board. Its profile is too limited for the broad discussion that needs to be had in the boardroom. Instead, look for someone who comes from a technology background, has run a technology-related business or company, understands other industries, and can apply what they know to other industries. We are looking for a former GM/Operator in a technology company who has Cyber/AI knowledge and has worked extensively as an Operator and GM. Avoid hiring CSOs, CIOs, and AI researchers/engineers unless you are an AI technology company. They would not have had the scope of knowledge to add value to the board on all the topics that the board needs to manage.
You have worked in a variety of sectors, including utilities, industrial technology, and water technology solutions. Are there any industries where AI-powered surveillance will have the biggest impact?
In the areas of back office and cyber automation, it can have a similar impact across the enterprise. What is different is the level of disruption they will cause to the industry based on their business model. Whether it's automating information, automating software design, or automating back-office workflows, it's arguably the fastest to transform service- and data-oriented industries. Consulting, customer service, technology, software development, advertising, and marketing companies are already being disrupted by this model. The financial sector has been a rapid adopter of AI due to its history of using technology to automate analytics and workflows, and the nature of its work that emphasizes research and analysis, not just workflows. Healthcare will see significant disruption in research and time-to-market for drug discovery and healthcare solutions.
What is one big risk that boards should be aware of when using AI?
Loss of proprietary data due to incorrect data, incorrect prompts leading to incorrect analysis or incorrect answers, or failure to use or implement a secure tenant LLM. Companies need to train everyone on how to use LLMS and how not to use LLMS: what data is acceptable and what tools to use. Like any other tool, data control governance needs to be in place for LLM.
Five years from now, what will separate the high-performing boards from the mediocre boards when it comes to how they use AI?
Literacy and expertise in the use of AI to innovate strategy and operations, mainstream application of AI where it is important and impactful, and a strategic committee to ask the right questions and drive the appropriate use of technology/AI. Boards must balance the risks of innovation and investment with the risks of inaction and delay.
