Best practices for employers using AI in employment law

With the use of AI increasing rapidly, it’s understandable that both employers and employees have concerns about its impact in the workplace.

According to ACAS, 26% of workers fear that AI could lead to job losses, while 17% are worried about it making errors. But what about employers? As AI continues to reshape how businesses operate, it is essential to consider the implications of AI in employment law. In this article, we explore what employers need to think about when implementing AI in their workplace practices.

AI presents enormous opportunities for businesses, but it also brings genuine challenges. Regulation in this area is extremely outdated and not fit for purpose. While there is currently no specific UK legislation governing AI, many obligations are still covered under existing laws such as the Data Protection Act, GDPR, the Human Rights Act, and the Equality Act.

This article is for information only and cannot be relied upon. You should always seek specialist legal advice where required.

Recruitment

Employers

It is a certainty that AI is already being used by some businesses to support decisions on hiring, and by candidates when applying.

However, AI tends to exhibit or even learn bias, in some cases despite safeguards being put in place to minimise the risk. This is a problem for an employer that relies on it for sifting through applications because if an individual is not selected as a result of biased AI output, it is the employer who will be liable for any subsequent discrimination claim. This is risky, costly and damaging to your reputation.

In order to reduce this risk, try to maintain human oversight and decision-making autonomy. You will also need to consider ongoing testing of the output of the AI algorithm to try to identify and remove any indications of bias.

Applicants

As for applicants who “enhance” their CV by using AI, you may want to consider having a disclaimer stating that an offer can be revoked (or employment later terminated) if the individual is found to have made a false or misleading application. Use this cautiously where people may have special needs in the application process though, for example, if they are sight impaired or have a neuro-diverse condition.

Finally, make sure that applicants are aware if AI is being used and be clear on your data privacy notice.

Employee use

You are likely to need an AI usage policy so staff know whether they can use it, how much and for what purpose. AI tools vary significantly and you may choose to allow some usage, but prohibit others. Be clear on this.

You also need to be clear on what information can be used (e.g. customer data) and what safeguards are in place. Ensure staff know their responsibilities in this area and continually monitor.

Think about accuracy and quality of work if staff are allowed to use AI tools rather than perform the work themselves. Is this what you pay them for? What kind of service is the customer receiving?

Finally, consider how you will manage things like copyright and the risks of intellectual property theft.

Data protection and privacy

You will also need to ensure that your data protection policy and privacy notices cover the use of AI for those you employ.

You must ensure that any employee data you use to feed AI has been lawfully obtained and that the use of the data is permissible. Failure to do so could lead to fines for misuse.

Special rules will apply if you send data to any third parties for processing.

Job safety

Many people are concerned about the impact of AI on workers who might be replaced. A report published in February 2025 by the Institute for Public Policy Research found that 70% of the tasks associated with knowledge roles could be significantly transformed or replaced by AI.

The potential for job automation in certain industries will mean that fewer employees will be required to carry out certain tasks. Think about whether you can upskill (and/or reskill) employees to adapt to AI and other technological changes.

This might form part of your longer-term succession planning.

Responsible use of AI

There is currently very little regulation around the use of AI and so, in the absence of regulation, employers will need to develop their principles for how and when they will use it.

Our tips are to ensure that your policies cover the use of AI, when and how it will be used and what controls are in place.

Run a thorough risk assessment, you will sometimes need to make value-based decisions on whether it is appropriate to use AI for a certain task. How reliable is it? What will happen if AI gets it wrong? What will the damage be and how will you deal with it? This requires really careful risk assessment. For data protection risk, the ICO has an AI and data protection toolkit that you might find useful.

Consider also whether you allow staff to use AI, is this them being efficient or is it being lazy? Are you aware that they are doing it, is the customer aware? Is the information being relayed even accurate and what checks and balances do you have in place? Consider overall transparency and accountability and how often you provide staff training on the safe use of AI.