At boardroom level, the stakes are high. Could AI ever function as a member of a corporate board?

“Unless board members understand generative AI and its implications, they will be unable to judge the likely impact of a company’s generative AI strategy and the related decisions regarding investments, risk, talent, technology, and more on the organisation and its stakeholders.” Four essential questions for boards to ask about generative AI, McKinsey

Artificial intelligence as a key player in the workplace is well and truly here. And yet, a Deloitte Global survey in 2024 found that nearly half (45%) of C-level respondents said AI hadn’t made it onto their board’s agenda at all – and only 14% reported discussing AI at every meeting

The wide-ranging survey (nearly 500 board members and C-suite executives across 57 countries) also found that many respondents knew that their board’s current level of engagement was not enough to deal with both the opportunities and the risks of using AI, particularly generative AI (Gen AI). Most respondents did not believe their organisation was ready for broader AI deployment. Only 3% of respondents thought their organisations were “very ready”, while 41% said their organisations were not ready.

Yet, the same survey found that boards are eager to devote more time to AI-related discussions. Nearly half (46%) said they were concerned that insufficient time was allocated to AI discussions. They are right to be concerned. Gen AI has the potential to transform many workplace processes, from the boardroom table outwards.

That’s something that CEOs are all too aware of, according to a new study by SAP. About 74% of executives are more confident in the business advice they get from AI than colleagues or friends.

Should boards be following their lead? Could AI, with its analytical capabilities, function as a member of a corporate board?

In fact, there are examples of boards which have been using AI successfully.

AI-powered boards of directors: Almost a decade in practice

In 2017, Deep Knowledge Ventures, a Hong Kong venture capital fund, added an artificial intelligence system for biotech investment analysis, enabling the identification of more than 50 parameters critical to assessing risk. The system, known as known as Vital, functioned as a “team member” on the board, and is credited with pulling the business back from the brink of bankruptcy. Vital, which stands for Validating Investment Tool for Advancing Life Sciences, helped the board make more logical decisions and helped them avoid investing in overhyped projects.

And in 2018 Salesforce CEO Mark Benioff told a World Economic Forum panel that the company’s artificial intelligence programme Einstein sat in on executive meetings every Monday morning.

In February 2024, the Abu Dhabi-based International Holding Company (IHC) reported that it had created a new non-voting board observer position for an AI agent called Aiden Insight. The virtual entity has the ability to process and analyse decades of business data, financial information, market trends, and global economic indicators, IHC says.

In Australia, the Real Estate Institute of NSW (REINSW) appointed an AI bot called Alice Ing as an adviser to the board. The Institute said in a May 2024 press release that Alice has an IQ of 155 and possesses an encyclopaedic knowledge of real estate.

We spoke to two AltoPartners experts about the potential role of AI in the boardroom, examining what boards need to do to use AI effectively and the crucial considerations they need to take into account.

Hesitation around AI in the boardroom is real

Kunal Ghosh, Director and Head of Engagement Accord India / AltoPartners India, notes an abundance of caution among board members.

On the most practical level, he says, many board members are worried about the use of AI in creating board packs. “They aren’t sure of the source of the data; they want the company management to qualify upfront if any part of what they’re seeing in a board pack has been derived or inferred using AI. That’s because they cannot rely on AI, given its current flaws and tendency to hallucinate, as the basis for approving a pack that is submittable to regulators, stock exchanges, or governments.

Boards are also worried about data risk. Kunal says there are concerns about the use of generative AI tools at board level. “If ChatGPT or any kind of generative AI is positioned at the board level and output from these tools is somehow leaked, that constitutes a significant data risk with significant potential penalties in terms of disclosures.”

Hesitation abounds around the regulatory environment too. “Country by country, people are waiting to see what AI regulations might come into play. The European Union has issued regulations. We know the United States is further along than most other countries. But generally, I don’t see a lot of countries rushing to put together a framework.”

Board members are not even sure they should use AI in their own professional capacities. Ghosh says a dipstick survey asked company board members if they had used AI as a professional board director. “The response was less than a percentage point – it was statistically insignificant. It seems no one is doing this.”

Carla Logan, Co-Head, Board Advisory and Governance Practice, DSG Global / AltoPartners USA, agrees and says the age of the average board member could be a factor. “Most board directors have tech-savvy leaders in their companies, but I don’t think there’s a lot of adoption on a personal or professional level.”

Logan also notes the importance of ethical considerations and employee well-being. “You’ve got to think about all the ethical considerations around AI. It would be very important for directors to understand how to interpret the data, all the outcomes they can get, and how those might potentially affect employees.”

Then there’s the sunk investment to consider. For companies who have already invested money in data science in their organisations, there are questions to be asked about whether investing in AI is worth it, says Ghosh. “If the output from AI is better than the output from data science, I would naturally gravitate towards AI – it’s simpler, easier to understand, instinctive, intuitive. But we found that companies are reluctant because they’ve already spent so much money on data science.”

What could boards be using AI for?

If boards were able to deal effectively with their very valid concerns, here’s how AI could be helpful.

1. Improving access to company information: AI platforms trained on a company’s intellectual property could allow board members to access relevant information from various internal sources.

2. Post-mortem analysis: Boards could apply AI tools to past data and decisions to understand if better outcomes could have been achieved, with potential applications in areas like supply chain and procurement, says Ghosh, who notes that these kinds of experiments are already happening.

3. Streamlining governance and compliance: AI can automate tasks such as risk assessment, legal document review, and data privacy compliance.

4. Providing comprehensive data-backed insights: AI can analyse large datasets from various sources, including market trends, customer behaviour, and financial indicators, to provide boards with comprehensive and real-time intelligence for better decision-making.

5. Enhancing collaboration: AI-powered tools – think virtual boardroom assistants and chatbots – could facilitate communication, information exchange, and sharing of insights among directors, regardless of their location or time constraints.

6. Supporting board succession planning: AI could potentially assist in succession planning, says Logan. “AI can help board members think about the gaps in their planning and analyse the possibilities over the medium term.”

7. Surfacing critical questions and alternative viewpoints: Boards could potentially use AI to generate additional critical questions on strategic issues or to provide an additional point of view to consider when making decisions.

What boards need to do to leverage AI effectively

The first step would be for board members to become more “tech-savvy” and develop a fundamental understanding of AI. “Everyone might need to go to ‘AI school’,” says Logan. “There’s a strong need for a real understanding of all emerging technologies so that boards can put effective policies in place.”

In addition to the knowledge needed to set policies and guidelines for AI usage within the organisation, board members could also think about including these areas in their “AI curriculum”:

1. Fundamental understanding of AI and generative AI – what these technologies are, how they work at a high level, their potential capabilities and limitations, the importance of human oversight and the balance between AI and human judgement.

2. How to interpret AI-driven data and outcomes – understanding how to deal with inaccuracy and hallucinations.

3. Ethical considerations – the possibility of biases in algorithms and the societal impact of AI deployment.

4. Risk assessment – cybersecurity threats, data privacy and breaches.

5. Data management and infrastructure requirements – the importance of robust data management systems.

6. Strategic implications – understanding where and how AI can be strategically applied to create value, improve efficiency, and drive innovation. This includes understanding how AI might affect the company’s industry, competitive position, and long-term strategy.

7. The evolving regulatory landscape – boards need to understand the rapidly developing AI regulations locally and globally and their implications for the organisation.

8. Talent and skills considerations – understanding the skills needed within the organisation and on the board to navigate the AI landscape.

How to start: caution, governance, leadership

Both Ghosh and Logan counsel companies and boards to move slowly, despite the hype around the urgency to adopt AI.

First off, Ghosh advises company leadership to think carefully about engaging consultants. “A lot of the consultancy firms are setting up their own internal AI units. That’s probably because clients are saying: ‘You are telling us to do AI, what have you done, other than a few reports and a few studies?’ As a director, you want to ask if you can offer your problems to AI. And then ask if you need to pay a consultancy to give you the same answer.”

He also wonders about the wisdom of creating a board role for somebody from an AI background. “If the person is biased towards ChatGPT, for instance, how does one mitigate that bias? I think most companies are thinking that over three to five years, there will be AI companies who emerge as winners, and there will be people left behind. From a legal and a tech framework standpoint, it makes sense to wait until you are dealing with one or two entities rather than an ecosystem with many players.” He also notes that it will only be clear over the next few years how the incremental costs of AI will stack up, and what the returns are.

On the question of governance, Ghosh foresees companies moving to advisory- or supervisory-level board initiatives “This could be a set of experts who have nothing to do with the company on a day-to-day basis, or even from a regulatory standpoint, but who work to understand real world problems and come up with real world solutions. Companies are saying this might be a better methodology than having the whole board take up the issue. AI vendors are pushing for AI to be on the board agenda because they know that that’s the way to create large scale adoption. Instead, this needs to be on the board agenda only after a few other people have gotten together and really thrashed it out.”

Logan advises starting slowly, with “low-risk” efforts based on a policy that can evolve over time. Those efforts, however, should be aligned with the company’s overall strategic objectives and be able to demonstrate shareholder value. For that to happen, executive management should take the lead and then filter their thinking to the board. “It’ll take a lot of leadership from the board chairs and the CEOs who are on those boards. It’s going to require each board director to get smart around AI.”

FURTHER READING

Ask Alto: Will AI be the best HR employee you hire in 2024?

AI Terms Every Leader Needs to Know

Ask Alto: How to make AI work in your organisation

Ask Alto: What is strategic empathy, and why is it a key leadership skill in 2025?

SOURCES

Generate Value From GenAI With ‘Small t’ Transformations

AI In The Boardroom: The Inevitable Evolution Of Decision-Making

Four essential questions for boards to ask about generative AI

Deloitte Global survey in 2024

5 Essential Elements of an AI-Ready Corporate Culture

Exclusive: CEOs are turning to AI for business advice and they trust it even more than their friends and peers (paywalled article)