Artificial intelligence is becoming a defining governance issue for boards globally. While much of the public discussion continues to focus on technology, the more important question for boards is one of oversight.

In our company’s conversations with board chairs and nomination committees, we increasingly encounter a common concern. Few non-executive directors (NEDs) doubt that AI is influencing their organisation. What they are less certain about is whether their board possesses the capability and governance structures required to oversee it effectively.

That concern is understandable. Many of today’s boards were assembled before AI emerged as a strategic priority. NEDs are now being asked to oversee technologies capable of influencing decision-making, customer engagement, workforce planning, operational resilience and long-term competitive advantage. At the same time, regulatory expectations are evolving through developments such as the EU AI Act and the UK’s emerging AI governance framework.

The challenge is not whether boards should engage with AI. It is how they can do so effectively.

Boards need judgement, not AI experts

One misconception encountered regularly is the belief that boards need to appoint a technology expert to solve the AI challenge.

In practice, effective governance rarely depends on a single specialist. Boards are collective decision-making bodies. Their effectiveness depends less on individual expertise and more on their ability to ask informed questions, challenge assumptions and exercise sound judgement.

Some of the strongest boards we work with at Principal Connections do not include recognised AI experts. What they do possess is intellectual curiosity, a willingness to learn and a culture that encourages constructive challenge.
Conversely, I have seen boards place excessive reliance on one technically knowledgeable director. While valuable, this can create a false sense of assurance. Other board members may become reluctant to engage with technology discussions, assuming the expert has matters covered.

AI governance should not sit with one director. It should become part of the board’s collective capability.

How should boards respond now?

The practical implications for NEDs and board chairs are becoming increasingly clear.

1. Establish accountability before problems emerge. One of the first questions boards should ask is simple: who is accountable for AI within the organisation? In many organisations, responsibility is spread across technology, legal, risk, compliance and operational teams. While this may reflect how AI is deployed, it can create ambiguity when governance issues arise. Boards should ensure ownership is clearly defined before problems occur rather than afterwards.

2. Focus on where AI is already being used. Many NEDs assume AI adoption is primarily driven by formal technology programmes. In reality, much of it happens organically. Marketing teams use generative AI to create content. HR teams use AI-assisted recruitment tools. Customer service teams introduce chatbots. Individual employees experiment with publicly available platforms.

The question we increasingly encourage boards to ask is not “Are we using AI?” but rather “Where are we using AI without fully understanding the implications?” Boards cannot oversee what they cannot see.

3. Treat data governance as a board issue. In discussions about AI, attention often focuses on algorithms. In my view, boards should be equally concerned about data. The quality of AI outputs will only ever reflect the quality of the data that underpins them. Poor data governance, biased datasets or weak controls can quickly become strategic, operational or reputational risks. NEDs should understand how data quality is monitored, how bias is assessed and what safeguards exist before AI-generated outputs influence significant business decisions.

4. Consider the broader organisational impact. The boardroom conversation about AI is often framed as a technology discussion. It is not. AI raises questions about workforce capability, organisational culture, customer trust, ethics, reputation and resilience. These are all areas that already sit within the board’s oversight responsibilities. The most effective boards are therefore considering AI through an enterprise-wide lens rather than viewing it solely as a technology initiative.

Board recruitment

Perhaps the most significant impact is in board recruitment itself. Over the past 12 months, we have seen a notable increase in board briefs seeking NEDs with experience in digital transformation, technology oversight or AI governance.

Historically, board appointments were heavily weighted towards sector expertise, financial leadership and governance experience. Those capabilities remain essential. However, nomination committees are increasingly asking an additional question: does the board possess the capability required to oversee a more digital and AI-enabled organisation?

The answer is not always to recruit a former chief technology officer or chief information officer. More often, organisations seek NEDs who have led transformation programmes, overseen technology-enabled growth, navigated regulatory complexity or successfully managed organisational change.

In some board searches, digital governance capability is becoming as important as traditional governance credentials.

The boards likely to succeed over the coming decade will not necessarily be those with the greatest concentration of technical expertise. They will be those capable of combining governance experience, commercial judgement and an informed understanding of how technology is reshaping their organisations.

Fast-evolving governance

AI is changing the questions boards need to ask. For NEDs, the challenge is not becoming technologists. It is ensuring that governance evolves as quickly as the technology itself.

The most effective boards are approaching AI in the same way they approach any significant governance issue. They are seeking visibility, demanding accountability, investing in capability and maintaining a healthy degree of constructive scepticism.

That approach is also influencing board recruitment. Increasingly, organisations are looking for NEDs who can contribute not only governance expertise and commercial insight, but also informed oversight of a more digital, data-driven future.

Ultimately, AI may be new. The principles of good governance are not.

Pat O’Donnell is managing partner at executive search consultancy Principal Connections