Generative AI and employment law: how to redesign work and keep employee trust.

This reflects Amara’s Law: we tend to overestimate the short-term impact of new technology and underestimate its long-term effects.

The history of other general-purpose technologies, such as electricity, suggests that transformational productivity gains do not come from simply adopting the technology itself. They arise when organisations redesign the way the technology is used.

A key starting point for employers is to analyse work at the task level rather than the job level.

Most roles consist of a combination of activities:

• Tasks that can be readily augmented through AI, such as drafting, translation, summarisation and information synthesis.
• Tasks that may be partially automated but require significant investment, oversight or specialist implementation, such as contract reviews.
• Tasks that remain fundamentally human, such as negotiation, relationship-building, leadership, trust-building and exercising professional judgement.

This means that organisations are increasingly asking not whether a role can be replaced, but which activities within that role can be redesigned.

For many employers, the objective is not to replace employees but:

• eliminating low-value administrative work;
• increasing productivity through AI-assisted drafting and analysis;
• reallocating employee time to higher-value activities; and
• redesigning teams, reporting structures and delivery models.

Perhaps the most reassuring message is that many of the capabilities employers value most remain difficult to automate. Trust, empathy, relationship management, communication, leadership and professional accountability continue to be fundamentally human attributes.

The future may therefore be less about replacing people and more about managing teams composed of both humans and AI systems.

Alongside recognising the continued importance of human skills, employers should take active steps to equip their workforce to work effectively with AI. This includes providing targeted training on how to use AI tools safely and responsibly, setting clear expectations around appropriate use (including confidentiality and data protection), and encouraging employees to develop “AI literacy” alongside core professional skills. In practice, this may involve introducing structured training programmes, creating internal guidance on acceptable use, and giving employees time to experiment with AI tools in a controlled environment. Employers who invest in both technical capability and human judgement will be better placed to realise the benefits of AI while maintaining trust and accountability.

The legal consequences arise when workforce redesign begins to alter employees’ roles in a meaningful way.

Employers may view AI as a productivity initiative. Employees may experience it as a fundamental change to their job.

Where AI adoption leads to significant changes in duties, responsibilities or reporting structures, employers should consider the legal challenges such as:

If a role no longer exists in its current form because certain activities have been automated or removed, a redundancy situation may arise. Employers cannot simply relabel a role while removing substantial elements of the work previously performed.

Where AI has reduced or removed the need for the original role, the employer may instead look for suitable alternative employment. Any alternative role should be a reasonable fit for the employee’s skills and experience, considering factors such as pay, location and status. Employees are generally entitled to a four-week trial period in the new role, and if the alternative is not suitable, they may be able to refuse it without losing their right to statutory redundancy pay. As with any significant change, this should form part of a proper consultation process.

Employers who dismiss employees for refusing to accept changes (or fire and rehire from January 2027) face potential unfair dismissal claims and, where 20 or more dismissals are proposed, collective consultation obligations. If an employer fails to follow a fair redundancy process, they face severe financial penalties. The primary consequences are a protective award of up to 180 days’ uncapped pay per affected employee for failing to consult, and from January 2027, uncapped tribunal compensation for unfair dismissal.

Experimentation and trial-and-error will be essential to successful AI adoption. From an employment law perspective, however, experimentation cannot come at the expense of following the correct process. Organisations introducing significant AI-related changes should consider early engagement with employees, staff representatives and trade unions where appropriate.

Many AI-driven transformation projects involve redesigning roles rather than eliminating them entirely.
Before implementing any changes, employers should consider whether they would affect the employment contract. If the proposed changes alter contractual terms, employers should first determine whether the contract already permits such changes. In some cases, the existing wording may be broad enough to cover a different way of carrying out the role, or there may be a flexibility clause dealing with duties, location or working arrangements. But these clauses are usually interpreted narrowly, especially where the change would disadvantage the employee or significantly alter the nature of the role.

Where the contract does not clearly permit the change, the safest course is usually to obtain the employee’s agreement. Imposing significant changes without consent can amount to a breach of contract. If the change is serious enough to undermine the employment relationship, it may also give rise to constructive dismissal claims. In practice, this means employers should approach AI-related role redesign through consultation and agreement, rather than assuming they can impose new duties or expectations unilaterally.

The organisations that benefit most from AI are likely to be those that redesign work most effectively. For employers, that redesign process creates both opportunities and legal risks.

AI is reshaping how work is done, but it has not rewritten the rules of employment law. Established legal principles continue to apply, even as the pace of change accelerates. For employers, the risk lies not in adopting AI, but in moving too quickly—overlooking process, bypassing consultation, or implementing changes without proper contractual footing.

The organisations that will benefit most are those that approach AI adoption strategically: redesigning roles thoughtfully, engaging with employees early, and ensuring legal compliance sits alongside innovation. Employers considering changes to contracts, roles or workforce structures should seek specialist employment law advice at an early stage, so that the process is properly managed and potential risks are identified and mitigated. With the right support, employers can embrace the opportunities AI presents while maintaining trust, reducing exposure to claims, and embedding change in a structured and sustainable way.