Hybrid governance is no longer an emergency workaround. Remote and hybrid board meetings have become part of normal corporate life, supported by board portals, video platforms and digital voting. Now artificial intelligence is beginning to affect how boards receive information, prepare for meetings and record decisions.
The move to hybrid working offers a useful lesson. Technology can change the mechanics of governance very quickly, but it does not automatically improve governance. A meeting can be easier to attend and still be poorly chaired. A board pack can be distributed securely and still obscure the decision required. A transcript can capture every word and still fail to produce useful minutes.
The same principle should shape the use of AI in corporate governance. AI can support better work, but only when the board is clear about the problem it is solving, the information it is using and the human judgement that remains accountable for the result.
Hybrid boards are different, not automatically better
Hybrid board meetings can widen participation. Directors can join without travel, urgent meetings are easier to arrange, and the chair can use digital functions to bring quieter voices into the discussion. Hybrid AGMs can also make participation possible for shareholders who cannot attend in person.
Those benefits come with practical trade-offs. A meeting that combines people in the room with people online is often harder to run well than a meeting in a single format. Remote participants can miss visual cues, informal discussion and the small interactions that help relationships develop. Newer governance professionals may also lose some of the learning that happens by observing experienced colleagues at close quarters.
The technology itself needs active governance. Access controls, meeting links, board portals and document permissions must protect confidential information. The contingency plan matters too. If the platform fails during a time-sensitive decision, the board needs to know how the meeting will continue and how participation and voting will be recorded.
The company secretary is central to making this work. Good hybrid practice requires more than sending a link. It requires deliberate agenda design, clear participation rules, reliable technical support and a chair who checks that everyone can contribute.
This is also why hybrid working should form part of a wider review of board effectiveness. The question is not simply whether a board can meet remotely. It is whether the format supports informed challenge, sound decisions and a reliable record.
Where AI can add practical value to governance
The most useful applications of AI are usually specific and unglamorous. Boards do not need to begin with a large transformation programme. They can start with defined tasks where the benefit, risk and quality of the output can be assessed.
Help directors work through board information
AI tools can help directors search long documents, identify recurring themes and create an initial summary of a board paper. They may also help a new non-executive director find previous discussions or understand how an issue has developed over time.
This could make board information more accessible, but it should not become an excuse for longer packs. If papers are reaching the board without effective committee scrutiny, clear executive summaries or a defined decision request, an AI summary treats the symptom rather than the cause.
Support onboarding and board development
AI can give directors a faster route into policies, previous decisions and technical terminology. It can also help governance teams tailor training to the questions a board is actually asking.
That support is particularly relevant as AI itself becomes a board topic. Directors do not need to become data scientists, but they do need enough literacy to challenge management on opportunity, risk, controls and assurance. In July 2026, research from The Chartered Governance Institute UK & Ireland found that half of governance professionals saw AI literacy for boards as an increasingly important responsibility.
Assist research and routine drafting
Governance teams may use AI to locate information, compare documents or produce a first draft. In subsidiary governance, for example, it can help a user work through information published in another language. It may also assist with the first stages of policy or briefing-note preparation.
The user still needs to know which question to ask, which source should be trusted and whether the answer is complete. AI can accelerate research. It cannot accept accountability for it.
Provide limited support around meeting records
Transcription and AI-assisted drafting may help governance professionals revisit parts of a discussion or prepare a rough first version. However, board and committee minutes are not transcripts. They need to capture the matters considered, the challenge offered, the decision reached and the actions agreed.
The board pack and the minutes together form the meeting record. Repeating every detail from the paper adds volume without clarity, while recording every spoken contribution can create a defensive document that is difficult to use. Experienced human judgement remains essential to decide what belongs in the final record and how it should be expressed.
The governance risks boards cannot delegate
The opportunity to save time is attractive, particularly for governance teams under pressure. The risks are equally practical.
Accuracy and invented information
Generative AI can produce a confident answer that is incomplete or wrong. It may rely on a weak source, miss a qualification or invent a supporting detail. Research published by The Chartered Governance Institute UK & Ireland in 2025 found that 74% of governance professionals were concerned about the accuracy of AI-generated content in corporate reporting.
Any material output therefore needs an identified human owner. Sources should be checked at origin, calculations verified and quotations compared with the underlying document. The higher the impact of the decision or disclosure, the stronger that review should be.
Confidentiality, data and cybersecurity
Board information can include market-sensitive data, personal information, legal advice and strategic plans. It should not be entered into an AI tool unless the organisation understands how the data will be processed, stored and accessed.
This is not simply an IT procurement question. Boards need assurance over approved tools, user permissions, retention, supplier terms, data location and incident response. The UK Government’s Cyber Governance Code of Practice places responsibility on boards and directors to govern cyber risk in the same way as other material business risks.
Loss of judgement and organisational voice
AI-generated writing can sound polished while flattening nuance. Board papers, minutes and governance advice need to reflect the organisation, the decision and the context. Generic language can conceal uncertainty or make weak thinking appear more settled than it is.
There is also a risk that convenience reduces scrutiny. If a summary becomes the only version a director reads, or an AI draft is accepted because it looks complete, the tool may narrow rather than improve the board’s understanding.
AI cannot fix poor governance design
Technology works best when the underlying governance is sound. It cannot compensate for unclear delegations, overloaded agendas, weak committee structures or papers that do not explain what the board is being asked to decide.
The same is true of hybrid meetings. A better camera will not create challenge. A secure board portal will not improve a badly written paper. Automated notes will not correct a meeting that reached no clear conclusion.
Before adopting another tool, boards and governance teams should ask:
- What decision or process are we trying to improve?
- Is the existing governance process working as intended?
- What information will the tool use, and is that use permitted?
- Who will review the output and remain accountable for it?
- What evidence will show that the change has improved quality, not only speed?
These questions keep technology connected to governance outcomes.
Seven practical steps for boards and governance teams
Define a small number of approved use cases
Start with tasks that have a clear owner and a reviewable output, such as searching approved materials, summarising a non-sensitive briefing or comparing document versions. Set boundaries around what the tool must not be used for.
Put policy around actual behaviour
Find out where AI is already being used. An effective policy should reflect real workflows, not an assumption that nobody has started. It should cover approved tools, prohibited data, human review, record-keeping and escalation.
Apply proportionate information security
Classify the information involved and match the control to the risk. Market-sensitive board papers require a different level of protection from a public training note. AI use should sit within the organisation’s wider risk management and compliance framework.
Keep a named human accountable
Every material output should have an owner who understands the subject and can defend the final result. “The system produced it” is not an acceptable explanation to a board, regulator, shareholder or court.
Require source verification
Where an AI tool supplies research, facts or quotations, users should return to the original source. The organisation should also be clear about when an output needs legal, regulatory, financial or technical review.
Build board and governance-team literacy
Training should help people make sound decisions about AI, not turn them into technical specialists. Directors and governance professionals need to understand the tool’s intended use, its limitations, the organisation’s controls and the questions they should ask.
Pilot, measure and review
Assess whether the tool improves accuracy, clarity, timeliness or decision support. Capture errors and near misses. Review the use case as the technology, regulation and organisation change.
The future of governance remains human
Hybrid working showed that governance can adapt quickly. It also showed that technology does not remove the need for preparation, relationships and judgement. AI will follow the same pattern.
The boards that benefit most will be curious without being credulous. They will test practical uses, protect sensitive information and insist that accountability stays with people. Governance professionals have an important role in connecting technology, decision-making and the evidence that good governance requires.
AI may change how the work is done. It does not change what the work is for: enabling the board to understand, challenge, decide and remain accountable.
If your board is reviewing its meeting practices, information flows or approach to technology, Beyond Governance can help. Get in touch to the team about governance that works in practice.
