Demystifying AI for the Board

To guide corporate AI strategy, boards can learn from the history of disruptive innovations and ground themselves in business value,  argues Tsvi Gal, CTO and Head of Enterprise Technology Services at Memorial Sloan Kettering Cancer Center.

Artificial Intelligence (AI) is inspiring both fascination and fear in boardrooms around the globe.

Directors are hearing that AI represents the next industrial revolution, and that the companies for whom they provide oversight must act fast or risk irrelevance. Yet at the same time, uncertainty persists among board members: What is AI really? How much of the hype is real? And how should boards engage responsibly to ensure that their respective companies unlock the potential value of AI while mitigating the significant risks that accompany these rapidly advancing capabilities?

The risks of failure are real. The IBM Watson for Oncology partnership with MD Anderson Cancer Center is a notable example. Despite ambitious claims about AI’s ability to recommend cancer treatments, the initiative was quietly dropped after spending over $60 million, as Watson struggled to interpret clinical data and expectations exceeded its technical capabilities.

The board’s role is not to become a team of technical experts, but rather to become fluent in the strategic questions that need to be answered. Those answers begin with gaining some perspective about the nature of AI: how it is both similar and different to past major innovations such as the internet. That context is something that we technology leaders are uniquely positioned to provide.

After all, this is not the first time a transformative technology has tested companies and their boards. From the long-ago introduction of electrification to the early days of the Internet to the advent of cloud computing, each technology-driven innovation has brought with it both disruption and opportunity. Boards that successfully navigated these shifts didn’t chase buzzwords. They committed to understanding the fundamentals (for example, what the cloud meant for their operations), asked better questions (What are our goals in deploying cloud? What is the ROI? What changes do we need to embrace to make it successful?). And they provided strategic clarity (as in “We plan to use cloud in these areas where it improves our productivity by X%, reduces our expanses by Y%, and shortens our time to market by Z%).

Taking a similar tack with AI will serve them well.

Pattern Recognition: Lessons from Past Transformations

AI may be advancing more rapidly than previous technological innovations. And the scope of its influence affects every facet of our lives. But, at its core, AI is ultimately no different than these previous shifts.

Therefore, history can serve as a useful guide. The Internet, the product of academic researchers and once relegated to IT departments, quickly became foundational to marketing, sales, operations, and product delivery when the World Wide Web emerged. Those companies that took too long to grasp its broader implications found themselves outpaced by more prescient and agile competitors. The same things happened with cloud and mobile computing. AI is following a similar arc, only things are moving even faster.

It's important to recognize the patterns that are repeating. Like its predecessors, AI is not a product; it is a general-purpose capability that can reshape how decisions are made, processes are optimized, and value is created. It has implications for every business function, from finance and HR to customer engagement and R&D.

For boards, the key fact is that AI is not just a tool but a force multiplier; it can unlock new efficiencies, reveal hidden risks, and enable faster, more strategic decisions. IT leaders can help their boards by demystifying AI’s capabilities, grounding expectations in practical use cases, and clarifying how AI aligns with core business goals and risk management.

Getting Beyond the AI Buzzwords

In order to become effective stewards of AI in the enterprise, board need to understand what AI is and what it isn’t. This is where IT leaders can help provide that much-needed perspective.

First, they can help board members develop a good working definition of AI. At its core, AI refers to systems that replicate or augment human intelligence. It includes tools as varied as machine learning models, natural language processing engines, and generative systems like ChatGPT, which allow users to get answers to questions using everyday language. AI may be able to learn from data, spot patterns, make decisions, or generate content. That breadth of capabilities and their potential business applications are what has generated such enthusiasm around AI.

Secondly, amid all this breathless excitement, IT leaders can urge their boards to take a dispassionate approach to AI. Specifically, there are three common traps that board members should avoid: 

1. Overestimating its capabilities. There is a tendency to think that AI can solve everything. In fact, it’s not always the right answer to a business problem.
2. Underestimating organizational readiness. AI demands a solid foundation of data governance, skilled talent, and new processes.
3. Treating it as a tech initiative. AI is not another IT project; it’s a business transformation enabler like the Internet, cloud, and mobile computing.

Finally, IT leaders can ensure that their boards understand that AI isn’t new. What is new is the democratization of AI. Thanks to cheaper compute, access to massive datasets, and — critically — the emergence of intuitive interfaces like ChatGPT, capabilities once blurred by technical complexity are now accessible to non-experts. This marks a fundamental shift. AI is no longer just the domain of data scientists or engineers; it is becoming a general-purpose tool that anyone can use.

What makes AI different from past innovations — and more urgent for boards to engage with — is that it doesn’t just change how we access or share information. It alters how decisions are made, how products are created, and how work itself is done. It introduces new capabilities and new risks at the same time, across every function. That duality requires not just curiosity, but strategic oversight.

From Uncertainty to Understanding: Educating the Board

An educated board is an empowered board. Technology executives can take the lead to help their boards gain a working understanding of how best to approach AI.

Some key lessons to impart as part of a deliberate AI education strategy include:

• Start with the business problem, not the tool. Technology leaders can help their boards anchor conversations in concrete business challenges — operational inefficiencies, customer churn, clinical outcomes — before introducing AI as a possible solution.
• Use historical analogies. Making comparisons to past transformative technologies can ease some anxiety around AI and illuminate strategic similarities and help to light a clearer path forward.
• Teach frameworks, not buzzwords. IT leaders can equip directors to ask a structured set of questions, such as: 
  o    What decision or process is being improved? This grounds the discussion in real business needs and helps distinguish hype from meaningful impact.
  o    Do we have the data, governance, and talent to support this? AI success depends not just on algorithms but on having clean, accessible data and the right people to manage it responsibly.
  o    What’s the cost of being wrong? Understanding the risk profile, whether reputational, financial, or regulatory, ensures the board can weigh AI investments against potential downsides.
  o    Who is accountable for the initiative and outcomes? Clear ownership is essential to avoid AI projects drifting without direction or measurable progress.
  o    How will we measure the ROI and manage unintended consequences? A disciplined approach to evaluation can help track whether the technology is delivering value and flag where it may be introducing new risks or inefficiencies.
• Bring the right voices into the room. Encourage board members to invite internal and external experts — such as the CIO or CTO; the Chief Data or AI Officer; external AI ethicists; risk professionals; or even operators leading real-world AI implementations in marketing, R&D, or customer service — to share their insight. Have them share real-world use cases as well as the operational challenges and strategic trade-offs involved in deploying AI.
• Integrate AI into governance structures. Consider creating an AI-specific subcommittee or integrating AI oversight into existing risk, ethics, and strategy committees. This ensures that AI initiatives receive the same level of scrutiny and accountability as other strategic initiatives.

This framework is just a start. Boards will also want to ensure they have strategic alignment, tying AI initiatives to business strategy; to oversee efforts to recruit the right AI talent to the enterprise; and to insist on robust governance to guard against bias, safeguard privacy, provide explainability, and comply with government rules. In addition, they should require clear metrics for returns on investment.

Forging Ahead, The Right Way

AI is not the future; it’s the present. Boards don’t need to be able to build models, but they do need to lead. Now is the time but not just to “do AI,” but to do “do AI right.” That means embracing a learning mindset, asking hard questions, and guiding management with informed oversight. Technology leaders can help them get there.

The stakes are high. The promise is real. And the winners will be those who move beyond fear and hype to build enduring capabilities, with clarity, discipline, and purpose.