Employment

The adoption and implementation of AI is likely to result in unprecedented efficiencies and cost savings for most employers; but it also raises a number of employment-related issues that need to be carefully considered and addressed.

AI is set to take over most repetitive, process-based tasks. This is leading to widespread concern in the workplace in relation to job security. Potential tensions are inevitable in circumstances where a highly efficient AI system is introduced which directly impacts upon existing employees and their roles.

This section discusses some of the employment law issues which may be relevant to organisations seeking to digitise their business by implementing AI.

Restructure & redundancies

AI, when adopted as part of a Digital Transformation strategy, may lead to a restructure within the organisation. Given the efficiencies offered by AI there is likely to be a reduced need for human employees. Although some employees may be reskilled into new and emerging roles, it is highly likely that any restructure will lead to redundancies. Depending on whether the employer intends to make fewer or more than 20 employees redundant, there are different consultation processes to follow which include minimum consultation periods.

Where the correct processes are not followed, there is potential exposure to unfair dismissal claims from employees for failure to carry out a fair redundancy process. There is the possibility of contract claims and claims for a protective award where the Employment Tribunal can award up to 90 days’ full pay per employee.

If the organisation transfers existing employees to newly created roles, the process will need to be carefully managed and handled correctly to ensure that constructive dismissal claims do not arise.

Where employers require new skills which are not available locally, talent will need to be recruited from overseas. This will give rise to immigration considerations such as whether a sponsor license needs to be obtained.

Where an employer has a unionised workforce, the unions will want to ensure that any collective bargaining arrangements cover the effect on its members of the introduction of AI into the workplace.

AI and bias

AI used to facilitate recruitment which is based on objective and well-designed criteria can be extremely beneficial for an organisation. By using unbiased AI it will be difficult for individuals to challenge recruitment decisions in the same way that they are able to do so when subjective decisions are made by human interviewers.

However, problems may arise where existing biases in the workforce are perpetuated. This can occur because AI recognises patterns and learns from historic data, meaning that any existing biases will be identified and reinforced in the AI’s decision making process. Under the Equality Act 2010, employers must not discriminate when recruiting staff, and where applicants are rejected during the recruitment process due to a protected characteristic (referred to in the diagram below) possible discrimination claims may result. Where a discrimination claim is successful in an Employment Tribunal, an employer’s liability is uncapped.

Policies and training

Employers should revisit their employment policies to ensure that they take into account their new, mixed workforce – with robots and AI working alongside employees.

In particular, employers also have wide-ranging health and safety responsibilities to their employees. Employers must ensure that their health and safety obligations and policies take into account the use of AI in the workplace because any breaches can carry hefty fines and criminal sanctions.

In addition, existing staff will need to be trained to work with, and alongside, the technology being introduced.

Current at 20 November 2020