Boardroom Essential

Why this matters at board level

When a board decides to part company with a CEO or senior executive whose conduct or performance has fallen short, the usual approach is to seek a swift exit via a settlement agreement. This may include paying an amount representing the maximum amount of unfair dismissal compensation (currently set at the lower of twelve months' pay or a statutory cap of £123,543). Since this cap is usually far exceeded by the executive’s annual remuneration package, there has to date been little incentive for executives to claim unfair dismissal. The existence of the cap has also made it relatively easy in most cases to reach a negotiated commercial settlement.

From January 2027, the financial risk profile of an executive dismissal will change fundamentally. A highly remunerated executive who brings a successful claim will be able to recover compensation reflecting their full loss of earnings, including not just base salary, but annual bonuses, long-term incentive plan awards, pension contributions, and other benefits. Losses can be claimed for whatever period a tribunal deems reasonable in order for the executive to find equivalent employment. For individuals at the top of the market, the sums involved can therefore be very large indeed.

Fair dismissal for executives?

One option for boards to consider is whether (and in what circumstances) it may be feasible to carry out a fair dismissal process for an executive. Performance management of very senior employees is rare and may be challenging, from a cultural and practical perspective. The employer (and to some extent the executive) would often prefer to act quickly, without disrupting the business and risking negative press coverage, shareholder concern or reputational damage.

Employers may be able to rely on “some other substantial reason” (SOSR), and this could allow for a simpler dismissal process. SOSR is one of the potentially fair reasons under the unfair dismissal legislation, alongside misconduct, lack of capability, redundancy or contravention of law. SOSR can be established where there has been a loss of confidence or breakdown in the relationship with the executive. But much will depend on the circumstances and the appetite within the organisation for taking this approach.

The settlement agreement lever

Settlement agreements are (and are likely to remain) the most effective tool for achieving a clean break in executive exits. In exchange for an agreed package, the departing individual waives their right to bring claims, including unfair dismissal. The challenge for boards is ensuring they retain sufficient leverage to secure that waiver, particularly where the individual knows that their potential claim is now uncapped. Valuation of losses from share and bonus awards may be complex, and lead to lengthy negotiations.

Tightening good leaver provisions

This is where remuneration committees should focus their attention. Many share plans and bonus schemes provide for "good leaver" treatment on termination, allowing leavers to retain the value of their awards. In many cases, good leaver status is either triggered by the particular circumstances of departure (e.g. redundancy or retirement) or determined by the committee at their discretion.

In light of the new uncapped regime, good leaver provisions should be reviewed. In particular, the remuneration committee should have express discretion to grant good leaver treatment only where the departing individual enters into a settlement agreement (if requested by the company) that includes a waiver of employment claims.

This creates a straightforward commercial dynamic: if the executive signs the settlement agreement, they receive the benefit of good leaver treatment across their incentive arrangements. If they do not, the committee retains discretion to apply bad leaver treatment, typically resulting in forfeiture of unvested awards and loss of bonus entitlements. This will remove much of the value from an uncapped unfair dismissal claim.

1. AI is tilting the cyber battlefield

Against the backdrop of headlines warning of AI models initiating cyber hacks, continued attacks on UK AI infrastructure, and the Trump administration banning access to some frontier models on national security grounds, the UK’s National Cyber Security Centre (NCSC) has been vocal about the ways in which AI is changing the cyber threat landscape. Its recent guidance on managing the cyber risk of agentic AI highlights the importance of carrying out threat modelling before deployment of agentic AI, controlling the agent’s environment with a robust sandbox, setting the right level of oversight and maintaining the ability to ‘pull the plug’ in an emergency. Boards should consider whether existing cyber risk frameworks adequately reflect the reality that AI is reducing the cost and increasing the scale of cyber attacks, while equivalent gains in cyber defence remain harder to realise.

2. Corporate governance rules in play

We continue to see increased focus, and investor scrutiny, on AI governance. The updated UK Corporate Governance Code also now requires boards to make an annual declaration on the effectiveness of their material internal controls (which may include technology issues like AI). For companies with calendar year-ends, the first declarations will be required in early 2027 – see our article on Provision 29 of the Code and the declaration of effectiveness in this edition for more information.

3. Increased AI reliance means increased operational resilience risk

As organisations embed AI into core processes, both internal and customer facing, operational resilience increasingly depends on the availability, performance and integrity of those AI systems. This is particularly true where critical services depend on a small number of AI model and system providers, creating potential concentration risks.  Boards need to satisfy themselves that AI dependencies have been mapped, stress-tested and factored into operational resilience planning. Cyber-style simulation exercises may also prove useful in rehearsing responses to AI-related disruption.

4. The EU AI Act is now generally applicable – and fines can follow

The AI Act has been in force for some time but with staged implementation. Most of the Act now applies, including the enforcement and transparency provisions which have applied since 2 August 2026, although the EU’s simplification programme for its digital and AI rules (the AI omnibus - see blog) has pushed back the timings for the Act’s high-risk obligations. The Act has wide extra-territorial reach (catching users with EU operations, but also user organisations outside the EU whose AI output is used in the EU) and fines are significant (up to 7% of worldwide annual turnover). Despite the Act's wide reach and significant penalties, many organisations are still being asked to make judgment calls in areas where guidance is limited – for example, around how the extra-territoriality works in practice, and even what constitutes an in-scope AI system (see our article for more information). Enforcement expectations also remain largely untested.

5. UK Government changes – what does this mean for AI Regulation

The arrival of a new prime minister has brought a reshuffling of AI policy priorities. The appointment of a new cabinet-level, dedicated Minister for AI and an AI Taskforce signal AI ambitions, while the dismantling of the Department of Science, Innovation and Technology (DSIT) has raised concerns about a lack of focus and loss of institutional expertise. No dedicated AI Bill has materialised, despite this being hinted at for a number of years, and a key question for the new administration will be how the UK wants to regulate AI risk - particularly given the fast paced evolution of the technology, and continued geopolitical pressures in this space. In the meantime, the UK’s sector-specific approach continues, meaning organisations need to continue engaging with relevant regulators and monitoring developments in this space. 

The 2018 Code of course already requires boards to monitor and annually review financial, operational and compliance controls, and to confirm in the annual report that a review had been carried out. As well as adding reporting controls to that list, the 2024 Code critically adds a new requirement that the board must in the annual report:

• describe how it has monitored and reviewed the effectiveness of the framework of material controls;
• make a declaration of effectiveness of the material controls as at the balance sheet date;
• describe any material controls that were not operating effectively at that date, together with action taken or proposed to improve them; and
• report on action taken in respect of any previously reported issues.

Importantly the declaration speaks as at the balance sheet date and the FRC has clarified in guidance that it does not expect companies to report on a material control failure from across the course of the year if it has been rectified by the balance sheet date (unless it was a publicised issue, as shareholders and stakeholders would already be aware of the issue and expect to see reporting on it).

2. What should reporting under Provision 29 look like?

It is up to each board to decide upon its material controls; there is no set number, but the FRC has indicated that many companies have identified around 30-50.  The annual report need not list those controls or describe the testing performed. Reporting should be proportionate and concise—normally no more than two pages—and avoid immaterial detail.

In terms of the declaration of effectiveness, there is no prescribed form of words. The board should give a clear, company-specific and reasonable conclusion, supported by a short account of its monitoring and review. A declaration may be qualified where a material control was ineffective. Any material weakness unresolved at the year end must be disclosed with a description of the actions taken or proposed to improve it but the FRC does not expect commercially sensitive or technical remediation detail that could harm the company’s interests to be disclosed. For illustrative purposes only, a ‘clean’ declaration might be along the following terms:

The Board has reviewed the effectiveness of the Group’s risk management and internal control framework and, having regard to the recommendation of the [Board Audit and Risk Committee] and its work throughout the year, has concluded that, as at [31 December 2026], the Group’s material controls were operating effectively, providing reasonable assurance against financial, reporting and operational losses and of compliance with laws, rules and regulations.

4. Provision 29 and the role of the auditor

The FRC has recently published a Provision 29 mythbuster that clarifies the role of the auditor in relation to Provision 29. It is important to note that the Provision 29 statement is not covered by the audit opinion. In other words, auditors are not required to state that in their opinion the Provision 29 statement provides a “true and fair view” or to give any other form of comfort on the statement.

Similarly, auditors are not required to assess which controls are “material” nor are they required to test the design, implementation or operating effectiveness of the material controls. The auditor may, however, choose to test the operating effectiveness of certain controls as part of their response to assessed risks in the audit. This may include controls which are also material controls under Provision 29.

For audit purposes, the Provision 29 statement sits within what is known under the relevant accounting standard as ‘other information’, meaning that the auditor must read it and consider whether there is material inconsistency between the statement and (a) the financial statements or (b) the auditor’s knowledge obtained during the audit.

Broadly speaking, therefore, the auditor must consider whether it has any information that casts serious doubt on the accuracy of the declaration. If the auditor considers that there is a material inconsistency, it must evaluate whether this gives rise to a material misstatement of the “other information”. If an unresolved material inconsistency remains, the auditor may be required to report it as a material misstatement of the other information. Conversely, if the board discloses an ineffective material control relevant to the audit, the auditor considers whether it is new information, reassesses the risks of material misstatement where appropriate and decides whether further audit procedures are needed.

In practice, auditors are likely to have significant knowledge of the company’s financial controls, but less knowledge about operational, compliance and other controls. In the course of their work, auditors will generally share their views on the strength and weaknesses of the internal control systems they encounter in their audit work with the audit committee and the board. The board can take some comfort if the auditors do not find any weaknesses, but must be mindful that the scope of the auditors’ review is limited. By way of example, the auditors will not tend to comment on operational risks that are outside their area of expertise if these do not impact the financial statements.

The proposed changes are aimed at individual and trust shareholders and they will need to be taken into account by companies when structuring any transaction involving such shareholders.

HMRC say that they are seeking to address inconsistencies where substantively similar payments receive different tax treatment. In general, capital treatment is more favourable for individuals and trusts than income tax treatment because the rate of tax and the amount brought into charge are lower.

Three areas are of particular relevance to listed companies.

1. Capital returns following holding company reorganisations

HMRC want to close what they see as a loophole that allows the amount of “good tax capital” on shares to be increased which enables returns of value to be made in capital form, rather than as income.

A common technique has been to insert a new holding company above the group. This has the effect of increasing the amount of "good tax capital" to the market value of the old parent company shares. HMRC now proposes to "freeze" the capital on shares in any new holding company at the amount originally subscribed for the underlying shares.

2. Demergers

Currently, the capital reduction route is often used by companies to carry out non-statutory demergers. The proposal to “freeze” the capital in shares (as mentioned above) will mean more demergers will be carried out through the statutory demerger route. HMRC is therefore proposing a number of broadly positive changes to the statutory demerger conditions to give more flexibility and provide simplification.

However, as a trade-off for these relaxations, HMRC proposes to remove the automatic right to appeal to the tax tribunal should a statutory demerger clearance be denied. Boards contemplating future separations will need to consider how the revised statutory conditions might apply.

3. Distributions from non-UK companies

Under current law, different rules apply depending on whether a distribution is made by a UK resident company or a non-UK resident company. In some cases distributions from non-UK resident companies are taxed as capital, whereas the same distribution from a UK resident company may be taxed, at least in part, as income.

HMRC proposes to align the two regimes by extending the full UK distribution rules to non-UK companies. Groups with offshore holding arrangements should consider whether this could affect returns to shareholders.

What is changing?

With effect from 6 April 2027, where scheme documentation does not allow an ongoing surplus to be refunded to sponsors, trustees will have the power to pass a resolution to permit it (subject to any conditions they determine). In addition, regulations will change the amount of surplus required before a refund can be made, potentially allowing release at lower funding levels.

Some practical considerations

Sponsors may want to start a conversation with trustees now about whether any surplus can be refunded to them and if so, on what terms. Any decision will need careful thought and to have regard to wide range of factors, including those set out below.

• Scheme objectives: The feasibility of release may depend on the scheme’s long term objectives and how close it is to buying-out liabilities with an insurer and winding-up. Where buy-out is already in contemplation, trustees may be more reluctant to run-on the scheme to allow for surplus release.
• Liabilities: Surpluses can be ephemeral and illusory. Trustees will need to consider how certain they are that all liabilities have been properly identified and costed for as there is a long history of schemes being required to pay out more than originally intended.
• Employer covenant: Trustees will need to consider the covenant strength of the sponsor and whether it would be able to support the scheme in the future should the funding position deteriorate following any refund of surplus. They may require additional financial information from the sponsor. The sponsor may also need to consider whether it is willing to put in place contingent support such as guarantees to reassure the trustees.
• Fiduciary duties: Trustees have fiduciary duties and must think about whether release of surplus is appropriate from a fiduciary perspective. Whilst there is no requirement to share any surplus between sponsors and members, the members’ interests are a factor trustees will need consider and in doing this, they should take into account issues such as the level of inflationary increases paid and whether benefits have previously been reduced. However, trustees should also consider the source of any surplus, which in many schemes is likely to be sponsor contributions.