Artificial intelligence (AI) is creating significant opportunities for businesses, but it also introduces legal, regulatory, operational and reputational risks.
As AI becomes increasingly embedded in business processes, boards of directors are expected to provide effective oversight and governance of its use.
Recent guidance and market commentary emphasise that AI governance is now a board-level issue requiring clear accountability, risk management and strategic oversight.
In this article, we explore the recent guidance and practical steps that boards can take to manage AI risks.
Boards do not need to become AI experts, but they do need to ensure that AI is subject to the same level of governance, oversight and challenge as any other strategic business risk. AI whilst useful, does not remove accountability. Human oversight is critical and AI should be used with caution where decision making is required as it cannot in practice make judgement calls and it cannot assume accountability. Those organisations that adopt a structured and responsible approach to AI governance will be best placed to realise the benefits of AI while minimising legal and regulatory exposure.
Under the Companies Act 2006, directors must, among other duties, promote the success of the company for the benefit of its members as a whole, exercise independent judgement and exercise reasonable care, skill and diligence. While AI does not create new directors' duties or alter existing duties, it changes how those duties must be discharged in practice. Boards should understand where and how AI is being used within the business, assess both its benefits and risks, and ensure appropriate governance frameworks are in place.
In practical terms, directors should:
Establish clear policies governing the use of AI, including clear objectives for it use, for example, enhancing accessibility, inclusion, and fairness in products or services.
Ensure appropriate oversight, monitoring and reporting mechanisms are in place, including, if appropriate, an independent committee to oversee the implementation and use of AI in the business.
Integrate AI risks into the company's wider risk management and compliance framework. AI is no longer the sole responsibility of the IT team, it must also be on the board's agenda.
Carry out impact and risk assessments. Regularly test and evaluate systems to ensure they align with the company’s policies and legal obligations.
Build oversight of AI into regular board and committee agendas through a combination of strategic reviews, periodic monitoring, and event-driven escalation covering its use, development and implementation. This could include areas such as compliance with, and updating, data protection policies to incorporate the use of AI including data retention periods and use of personal data, confidentiality, intellectual property, cybersecurity, regulatory requirements/expectations, and bias and discrimination.
Ensure sufficient human oversight of AI-generated outputs and decisions.
Invest in training so that directors and employees understand the opportunities and risks associated with AI.
Establish clear lines of responsibility. The board should be closely involved in, and retain ultimate decision-making, including veto powers over, AI implementation or continued use where significant commercial, regulatory, safety, or reputational risks exist.
Whilst AI can be used to record meetings and to analyse and summarise the content of such meetings, care should be taken to disable recording where there are sensitive discussions or discussions where the output is not intended to be shared, such as, discussions under Chatham House Rules. In addition, entering confidential information into AI tools may jeopardise privilege, so boards should consider disabling recording, transcription and AI for privileged discussions and keep privileged minutes clearly separated. All AI generated minutes must be carefully reviewed and verified for accuracy as they remain the definitive corporate record.
AI can be a real benefit to the business but failure to oversee AI properly can expose both the company and its directors to a range of consequences. These may include regulatory enforcement action, data protection fines, discrimination claims, intellectual property disputes, shareholder litigation, and significant reputational damage. Directors may also face scrutiny for breaching their statutory duties where they fail to properly identify, manage or supervise material AI risks affecting the business. Increasingly, regulators and courts are examining issues such as misleading claims about AI capabilities, algorithmic bias and unlawful use of personal data.
For further reading, ecoDa (the European Confederation of Directors’ Associations) and Accountancy Europe have jointly published a paper entitled 'AI Governance - Ten Principles for Effective Board Oversight (October 2026), available here (full access may require membership). The Chartered Governance Institute UK & Ireland has also produced a Guidance Note on Digital Boardroom Governance: Board Meeting Hygiene in the Digital and AI-Enabled Boardroom. A link to the note is here.
This article is intended as general guidance only and does not constitute legal advice. For more information, please contact our corporate team.