From AXA to Tesla, major companies are embracing Gen AI in a big way for various governance aspects, including board governance, recognising its potential to revolutionise various industries. Already it has transformed how organisations operate, communicate, and innovate.
Indeed, AI continues to enhance decision-making, governance, and operations, but replacing the human-driven boards of enterprises entirely with it poses major ethical and functional challenges. Chris Weitz’s latest thriller movie “Afraid”, about AI taking over and terrorising a household, may be an extreme possibility.
Remember also what the “Godfather of AI” and this year’s joint Nobel Prize winner in Physics for his work on machine language, Geoffrey Hinton had said after quitting from Google last year. Although he believes AI has immense potential for good, he is increasingly concerned about its risks, particularly as companies and governments continue to develop more advanced AI without fully understanding its implications. And boards adapting AI without considering this may be jeopardising the real intelligence.
Role of boards
Boards of directors have the fundamental role of governance and oversight. They provide strategic direction, ensure compliance with regulations, manage risks, oversee financial performance, and safeguard the interests of stakeholders. A board must offer a diverse mix of experiences, skills, and perspectives to provide guidance on complex issues, including business strategy, ethics, sustainability, and succession planning. Given these responsibilities, can AI realistically fulfil these roles, or at least supplement them?
Many organisations globally have allowed AI into their boardrooms and governance processes, providing valuable insights, automating tasks, and facilitating more efficient decision-making. There are a few areas where AI is showing its potential:
#JPMorgan Chase uses Gen-AI to provide data-driven reports for risk committee. They analyse operational data and identify emerging risks, which are presented to the board for oversight. The AI’s ability to detect patterns that might go unnoticed by human analysis has significantly improved the board's risk management capabilities Gen-AI excels at processing vast amounts of data in a short period, making it a valuable tool to make data-driven decisions. Boards can make better-informed strategic decisions for business growth.
#Ping An Insurance in China is using AI for compliance monitoring and risk assessment. Such tasks require precision and can be time-consuming. By automating them, AI can save boards significant time and resources, allowing board members to focus on more strategic matters.
Why human intelligence can’t be replaced
While AI’s contributions are clear, the prospect of completely replacing human boards with AI remains far-fetched for the time being.
Firstly, the most critical aspect of human judgement and ethics will be absent. Even generative AI currently lacks the human capability for judgment, ethical reasoning, and empathy. Boards deal with many ethical dilemmas, which require a deep understanding of societal norms, human emotions, and morality.
Take for instance the death of a young employee at EY India due to work pressure. The board of EY will have to work on multiple levels of this complex issue to ward of negative PR and also to create safer workplaces. With little or no data, AI will go totally haywire at such situations. AI certainly cannot account for ethical ambiguities. It can provide data-driven insights, but lacks the ability to weigh moral dilemmas and understand the ethical implications of its decisions.
Second, boards need strategic leadership and vision, which no Gen-AI can provide as of now. AI excels in operational tasks, but governance requires traits that are deeply human. Boards must make decisions based on incomplete data or intuition, something that AI, with its dependence on existing data sets, is not adept at. For instance, a board may need to decide whether to enter a new market or launch a ground-breaking product. These decisions are not purely data-driven; they require an understanding of the competitive advantages, value chains, customer behaviour, and broader economic conditions – all of which involve human intuition and creativity. AI can inform strategy, but it cannot lead it.
Third, one cannot expect AI to handle complex stakeholder relationships. Boards must balance the interests of multiple stakeholders – shareholders, employees, customers, suppliers and the broader community. AI lacks the interpersonal skills necessary to engage with and manage these complex relationships. The role of boards is not just about making decisions; it’s also about building trust, negotiating with various parties, and building a certain culture and values. When boards negotiate with unions over employee wages or benefits, human intuition, negotiation skills, and empathy are crucial, which AI cannot provide as of now.
Finally, and this is going to be a very critical issue in the years to come, is the safety and security of AI algorithms. AI is only as good as the data it’s trained on, and there’s a well-documented risk of AI models reflecting biases present in their data. This could lead to biased decision-making, especially if AI were to take on more autonomous roles in governance. When a board needs to consider diversity and inclusion when hiring executives, and if the AI is trained on biased data, it may inadvertently favour candidates from specific backgrounds, perpetuating existing inequalities.
Rather than a full replacement, the most realistic future is one of collaboration, where AI empowers boards to be more efficient, informed, and adaptive, but where the ultimate decisions still lie in human hands. When AI starts running the boardroom, just hope it doesn't schedule a meeting with Skynet, the self-aware AI from Terminator that views humanity as a threat!
Discover the latest Business News, Sensex, and Nifty updates. Obtain Personal Finance insights, tax queries, and expert opinions on Moneycontrol or download the Moneycontrol App to stay updated!
