Anurag Bana

Legal Policy & Research Unit, International Bar Association

Anurag.Bana@int-bar.org

Niharika Julka

Legal Policy & Research Unit, International Bar Association

julka.niharika9@gmail.com

Introduction

Seventy-five years ago, Professor Isaac Asimov published The Evitable Conflict, a short story depicting a future where autonomous machines managed global industries. For decades, that premise remained firmly in the realm of fiction, but modern corporate reality is gradually catching up to Asimov’s imagination. Artificial intelligence (AI) is now embedded in nearly every aspect of modern life and has become an integral part of how organisations operate, compete and make decisions.^[1]

For-profit boards are rapidly shifting from general exploration to making generative artificial intelligence (GenAI) an absolute strategic priority. Recent data indicates that over three-quarters of companies have actively deployed AI in at least one business area, and 70 per cent of them feel sufficiently knowledgeable to dictate corporate AI policy.^[2] The appetite for this integration continues to intensify. A 2024 Deloitte Global survey^[3] revealed that board members are eager to dedicate significantly more time to technology-focused discussions, with nearly half expressing concern that their current meeting agendas do not allocate enough time to these critical debates. This urgency stems from a stark reality, as AI is already actively transforming high-stakes decision-making by providing insights that shape everything from long-term strategy to daily risk management.

This transformation has been unfolding in phases around the globe for nearly a decade. As early as 2014, the Hong Kong venture capital fund Deep Knowledge Ventures appointed an AI system named Vital to analyse biotech investments.^[4] By evaluating over 50 critical risk parameters, the system helped steer the company away from financial collapse and earned actual voting rights on investment decisions.^[5] Later, in 2018, Salesforce CEO Marc Benioff revealed at the World Economic Forum that the company’s AI platform, Einstein, regularly attended executive meetings.^[6]

More recently, AI participation in boards has accelerated dramatically. In early 2024, the Abu Dhabi-based International Holding Company announced the appointment of ‘Aiden Insight’, an AI-powered non-voting board observer, capable of ‘continuous data analysis, risk assessment, strategic planning support, innovation tracking, and ethical and compliance monitoring’.^[7] That same year, Australia’s Real Estate Institute of New South Wales appointed a virtual adviser named Alice Ing to its board, leveraging her encyclopedic grasp of property data to eliminate blind spots.^[8]

Even traditional banking giants are embracing this shift, evidenced by Lloyds Banking Group becoming the first FTSE 100 company to officially integrate an AI system into its board operations.^[9] Although currently operating as a pilot programme,^[10] the initiative may establish an important precedent for large corporations worldwide.

Crucially, state level institutions are now adopting this technology as well. In 2025, Kazakhstan appointed AI as an independent board member with voting rights to its sovereign wealth fund, Samruk-Kazyna Fund.^[11] This indicates that government and public asset managers are mirroring the private sector’s aggressive push towards automated governance.

The increasing reliance on AI is also changing how executives perceive advice and expertise. According to a recent study by SAP,^[12] approximately 74 per cent of executives reported greater confidence in business advice generated by AI than advice received from colleagues or friends.

The emergence of AI in the boardroom raises profound legal, ethical and governance questions. Can AI exercise fiduciary judgment? Who bears responsibility when AI-assisted decisions cause harm? And how should corporate law adapt when machines begin influencing or potentially making strategic decisions traditionally reserved for human directors? These questions are no longer theoretical. They are rapidly becoming central issues for modern corporate governance.

Why AI matters in the boardroom: the emerging role of AI in corporate governance

At its core, corporate governance suffers from a structural vulnerability known as the agency problem.^[13] Because a board of directors must rely almost exclusively on the management team for operational data, market updates and performance metrics, they are inherently dependent on the very executives they are tasked with overseeing. This may inadvertently restrict a board’s independent judgment.^[14] Moreover, even though executive teams are generally diligent, the information they provide for meetings may lack external context or alternative strategic perspectives. AI can alter this dynamic at a fundamental level.

AI tools can enable directors to independently analyse public disclosures, market intelligence, industry benchmarks and macroeconomic trends in real time. Directors may also use AI to review and synthesise historical board materials, management reports and prior board packages, allowing them to identify long-term patterns, inconsistencies or overlooked risks that might otherwise remain buried in large volumes of information.^[15]

Practical experiments involving AI in executive decision-making have already demonstrated both the promise and limitations of these tools. In a study conducted with the Austrian company Giesswein, researchers explored the integration of AI into executive meetings to assess how it influenced boardroom dynamics.^[16] The experiment found that AI could enrich discussions by encouraging executives to consider alternatives and perspectives they might not otherwise have explored. Importantly, the technology’s greatest value was not necessarily the generation of answers, but its ability to interrupt entrenched patterns of thinking and slow down decision-making sufficiently to encourage deeper reflection.^[17] However, researchers also concluded that AI remains heavily dependent on active and thoughtful human engagement.^[18] Effective use of these systems requires critical thinkers capable of interrogating and contextualising AI-generated outputs rather than passively accepting them.

As boards become more comfortable with AI technologies and address legitimate concerns surrounding confidentiality, reliability, bias and accountability, AI may become increasingly embedded in governance processes. Other potential applications include the following:

• Post-decision analysis: Boards could use AI to analyse historical corporate decisions and outcomes to determine whether alternative strategies may have produced better results. Such retrospective analysis may be particularly valuable in areas such as supply chain management, procurement, investment strategy and operational risk assessment. These forms of AI-assisted governance experimentation are reportedly already being explored within industry.^[19]
• Data-driven strategic intelligence: AI systems can process and synthesise enormous datasets from internal and external sources, including financial metrics, customer behaviour, market trends, geopolitical developments and economic indicators.^[20] This capability may provide boards with more comprehensive, real-time intelligence to support strategic decision-making and oversight responsibilities.
• Enhancing board-management engagement: Beyond improving oversight, AI may also enhance the quality of strategic dialogue between boards and management.^[21] More sophisticated AI systems are capable of blending internal company data with external market intelligence to pressure-test strategic assumptions, identify emerging risks and evaluate alternative scenarios.^[22]
• Assessing board performance: AI can benchmark a board’s governance practices against peer organisations and recognised governance frameworks by analysing materials such as board agendas, committee charters, meeting minutes and skills matrices. This may allow directors to evaluate whether board time and expertise are being allocated effectively and whether governance structures align with leading practices.^[23]
• Predictive scenario planning: AI systems are increasingly capable of combining macroeconomic indicators with company-specific performance metrics to generate predictive forecasts and model best-case, worst-case and baseline scenarios.^[24] Such capabilities may help directors evaluate the implications of strategic decisions involving capital allocation, mergers and acquisitions, regulatory changes, cybersecurity threats and broader governance vulnerabilities.^[25]

AI, therefore, has the potential to strengthen both oversight and strategic engagement. However, its value ultimately depends on how it is integrated into board practice.

The legal boundaries of AI in the boardroom

As board members increasingly incorporate AI into boardroom practices to support their fiduciary duties, they are likely to raise concerns about its responsible use, along with the associated legal and broader risk implications.

A central legal constraint on the use of AI in corporate governance is that, in most jurisdictions, directors must be natural persons.^[26] Under the UK Companies Act 2006,^[27] for example, every company is required to have at least one director who is an individual. This foundational principle anchors corporate accountability in human decision-makers and limits the extent to which governance functions can be delegated to artificial systems.

Against this backdrop, the key legal question is not whether AI can participate in the boardroom, but how far its role can extend within the boundaries of lawful delegation.^[28] In practice, this involves distinguishing between decision-making authority and decision support. While AI systems may assist in analysing data, generating insights and identifying risks, ultimate authority must remain with human board members who are bound by the duty of care, as well as the business judgment rule in commercial decision-making contexts.^[29]

Some legal scholars argue that boards should treat AI outputs similarly to the advice provided by external human experts.^[30] However, the board must appropriately account for the structural differences inherent in algorithmic logic versus human expertise. Human board members must retain ultimate decision-making power and cannot blindly rely on machine recommendations to shield themselves from claims of negligence or failure to exercise independent business judgment.

Most legal frameworks therefore do not permit the transfer of core management authority to AI systems.^[31] Its permissible roles are generally limited to advisory or observational functions, such as providing analytical support, generating forecasts or assisting with the evaluation of complex datasets. However, the law may evolve in response to further technological developments.

Crucially, these limits on delegation do not mean an AI system must be fully explainable or legally liable on its own before a board can utilise it. Instead, responsibility remains with the board members who deploy and rely on such systems.^[32] They are expected to discharge their duties of care in overseeing how AI tools are selected, trained, validated and deployed, as well as in interpreting and challenging their outputs.^[33] This requires a meaningful level of AI literacy,^[34] as highlighted under the European Union Artificial Intelligence Act.^[35]

Ultimately, fiduciary duties remain non-delegable. Duties of care and loyalty continue to rest exclusively with human directors, regardless of the extent to which AI systems inform or shape board deliberations.^>[36] This means that while AI may increasingly function as an advanced analytical partner, it cannot displace the legal and ethical accountability that defines the board members.

Moving forward: what boards must do to leverage AI responsibly and effectively

As AI becomes increasingly embedded in corporate governance processes, boards are no longer asking whether AI will affect boardrooms, but how it can be integrated responsibly, effectively and lawfully.

The International Bar Association’s (IBA) ‘The Future is Now: Artificial Intelligence and the Legal Profession’ report identified data governance, cybersecurity, intellectual property and privacy as among the most significant challenges associated with AI adoption, regardless of organisational size.^[37] These same concerns also arise in the boardroom context, where discussions routinely involve highly sensitive information, including strategic plans, financial forecasts, acquisitions, litigation and regulatory matters. The use of GenAI tools raises legitimate concerns regarding how such information is stored, processed and protected.^[38] Consequently, many companies often implement software guardrails limiting access to confidential or sensitive information to prevent unauthorised disclosure. This proactive stance aligns directly with broader risk oversight mandates. In this regard, the IBA’s report, ‘Global perspectives on protecting against cyber risks: best governance practices for senior executives and boards of directors’, emphasises that boards and senior management must understand both regulatory expectations and their organisation’s strategy for protecting critical systems and data.^[39] Here, effective oversight requires awareness of risk mitigation measures, technological developments and emerging threats, enabling directors to make informed decisions about the governance and deployment of AI within their organisations.^[40]

Boards are also operating within an uncertain and rapidly evolving regulatory environment. Different jurisdictions are adopting different approaches to AI governance, creating uncertainty for multinational organisations.^[41] The EU has already introduced comprehensive AI regulation through the European Union Artificial Intelligence Act, while other jurisdictions continue to develop or debate their own frameworks.^[42] This fragmented regulatory landscape makes it difficult for boards to establish long-term governance strategies with confidence.

In addition to legal uncertainty, boards must also consider broader ethical implications. Directors increasingly need to understand not only what AI systems can do, but how AI-generated outcomes may affect employees, workplace culture, decision-making fairness and stakeholder trust.^[43] Against this backdrop, boards seeking to leverage AI effectively will need to take the following measures:

• Developing AI literacy at board level: Board members must develop a foundational understanding of AI, including how the technologies operate, their capabilities and limitations and the risks associated with their use.^[44] Boards should establish formal AI literacy and education programmes to ensure board members are capable of critically assessing AI-generated output rather than completely relying on it.
• Establishing a culture of responsible innovation: Employees should feel able to explore AI tools openly and seek guidance without fear of stigma or organisational resistance.^[45]
• Strengthening oversight and accountability structures: Responsibility for AI oversight should be assigned within the organisation, potentially through a dedicated committee, designated executives or the appointment of a Chief AI Officer.^[46] Governance frameworks should also establish protocols regarding how AI outputs are reviewed, challenged, documented and incorporated into decision-making processes.
• Preparing for greater investor scrutiny: As AI becomes more integrated into governance processes, investors and other stakeholders are likely to demand greater transparency regarding how AI is being used in strategic oversight and decision-making.^[47] Boards should therefore be prepared to explain their AI governance practices, including the safeguards and accountability mechanisms surrounding AI-assisted decisions.
• Treating AI governance as an ongoing process: AI integration should not be treated as a one-time initiative. Boards will need to continuously reassess their AI practices as technologies, risks, legal obligations and market expectations evolve.^[48] Periodic reviews, continuous improvement mechanisms and ongoing monitoring will be essential to ensuring that AI governance remains effective and aligned with organisational objectives.

The growing importance of AI governance also highlights the need for stronger collaboration between boards and legal advisers. A recent survey^[49] of public company corporate directors suggests that many boards currently receive limited assistance from general counsel on AI-related issues, despite the significant legal and regulatory implications involved.

The legal profession and the future of AI governance

The growing use of AI in boards also raises broader questions for the legal profession itself. As boards increasingly look to legal advisers for guidance on AI-related risks, regulatory compliance, cybersecurity, governance frameworks and ethical oversight, there is a corresponding need for lawyers to develop sufficient understanding of AI technologies and their implications.

The IBA, through its global network of lawyers, in-house counsel, law firms and bar associations, is uniquely positioned to lead international collaboration and knowledge sharing on AI governance. By facilitating the exchange of best practices, educational resources, practical guidance and emerging regulatory developments, it can help shape and strengthen AI literacy across the profession and support lawyers advising boards in increasingly complex governance environments.

The emergence of differing national and regional AI frameworks presents significant challenges for multinational organisations and the lawyers who advise them. Working alongside regulators, policymakers, industry stakeholders and national bar associations, the IBA can advance efforts aimed at reducing unnecessary regulatory fragmentation, encouraging greater harmonisation where appropriate and supporting the development of uniform governance frameworks capable of evolving alongside technological change while preserving fundamental legal and ethical principles.

Equally important is ensuring that the legal profession has a meaningful voice in the development of future AI governance frameworks. The IBA is well placed to advocate for comprehensive and multidisciplinary consultation processes that bring together regulators, legal professionals, technology experts, industry representatives, academics, civil society organisations and end users to ensure that AI regulation is balanced and informed by a broad range of perspectives.

The legal profession also has an important role to play in the development of global best practices for AI governance. This may include guidance on issues such as data governance, information security, intellectual property protection, confidentiality, privacy, accountability and responsible AI deployment. By supporting the development of practical governance frameworks and encouraging organisations and law firms to adopt clear AI policies, the profession can help ensure that AI is deployed in a manner consistent with the rule of law, professional obligations and public trust.

Conclusion

AI is rapidly transforming corporate governance. What was once considered a futuristic concept is increasingly becoming a practical reality within boardrooms around the world. From risk management to predictive planning, AI systems are beginning to influence how directors access information, evaluate decisions and engage with management.

Current legal frameworks remain firmly grounded in the principle of human responsibility. Directors cannot delegate their fiduciary duties to AI systems, and ultimate authority continues to rest with human decision-makers. AI may support governance, but it cannot replace the judgement, ethical reasoning and accountability expected of directors and board members under corporate law.

Although most organisations are not yet ready to formally appoint an AI system to their board, growing pressure on decision-makers and those providing critical information is driving the search for new competitive advantages, making such developments increasingly plausible in the future.^[50]

While the rise of AI in governance presents significant opportunities, it also exposes organisations to new legal, ethical and operational risks involving accountability, bias, confidentiality, transparency and overreliance on automated systems. The challenge for modern boards, therefore, is not whether to engage with AI, but how to do so responsibly. Boards will need to invest in AI literacy, maintain human oversight, establish clear accountability structures and remain responsive to evolving legal and regulatory standards.

The legal profession will play a central role in this transition. As AI becomes increasingly embedded in corporate governance structures, lawyers will be expected not only to advise on compliance and risk, but also to help boards understand the broader governance implications of AI-assisted decision-making. Meeting this challenge may require greater collaboration across the profession, including the development of shared educational resources, best-practice guidance and cross-border dialogue on AI governance. International organisations such as the IBA can facilitate these efforts, support the development of globally informed governance standards, and contribute to the development of AI governance frameworks that are effective and consistent with the principles of accountability, transparency and the rule of law.

Looking ahead, AI capabilities will continue to evolve, and corporate governance is likely to undergo significant structural change. The most effective boards will not be those that resist technological transformation, nor those that completely surrender decision-making to AI, but those that successfully integrate AI into governance while preserving human judgment, accountability and ethical responsibility. In the years ahead, the ability to strike this balance may become one of the defining characteristics of effective corporate leadership.

Notes