The introduction of, and increasing reliance on, Artificial Intelligence (“AI”) in the workplace brings great advantages to businesses, but it also has potential complications and drawbacks for legal advisers, including risks to the unlawful dissemination of personal data, breaches of confidentiality, and potential breaches of equality laws.
What is AI?
Artificial Intelligence (or “AI”) is technology that attempts to mimic human intelligence by using, and trying to learn from, data provided to it, and adapting its behaviour to improve its performance in tasks normally undertaken by humans. AI applications include, among others, Chat GPT, virtual assistants, chatbots, and text analysis tools.
How is AI used in the workplace?
AI can be used in a number of different ways in the workplace, including (but not limited to):
- Recruiting and hiring (identifying the skills needed for a role; writing job descriptions; finding candidates; screening candidates; scheduling interviews; ranking candidates; and undertaking interviews)
- Performance management (allocating tasks; measuring the performance of individuals; analysing productivity; analysing attendance; and rating productivity)
- Managing remote workers (monitoring attendance; monitoring productivity; and determining when employees are starting and finishing their working days)
- Automation (automating repetitive tasks; undertaking clerical tasks; and increasing efficiency in the workplace)
How does UK law govern the use of AI in the workplace?
There is no specific UK legislation that directly governs the use of AI in the workplace. However, there is legislation which indirectly governs AI, including (but not limited to):
- Data protection laws, such as the UK General Data Protection Regulations (“GDPR”)
- Employment law, such as the Equality Act 2010 (which, in general, covers discrimination) and Employment Rights Act 1996 (which, among other things, governs the processes that should be used when dismissing employees)
- Intellectual property laws (particularly the implications for copyright law)
- Health and safety laws (mainly under the Health and Safety At Work etc Act 1974)
The UK is also a signatory to an international AI treaty named the Council of Europe Framework Convention on Artificial Intelligence, Human Rights, and the Rule of Law – new domestic legislation is expected in this Parliament in order to give effect to the provisions set out in the Convention.
What must legal advisers consider when dealing with AI in the workplace?
Legal advisers should generally consider the following when thinking about how AI is implemented in the workplace:
- The impact on data protection for the personal data of employees and third parties
- Whether the use of AI could lead to unintentional breaches of confidentiality regarding sensitive business information
- Whether the use of AI could lead to discrimination occurring in the workplace
- Whether the use of AI could lead to breaches of an employee’s contract, resulting in potential constructive dismissal claims
Data protection
The use of AI applications by employers is likely to mean that the personal data of employees and third parties (e.g. customers, clients, suppliers etc.) is processed by these applications. The processing of the personal data of data subjects must be done in a way which is compliant with the data protection principles outlined in the UK GDPR. Lawyers must therefore be careful to understand whose personal data is being processed by AI applications, how it is being processed, why it is being used, and where it is being processed in order to ensure that UK data protection legislation is complied with.
Confidentiality
The use of AI applications in the workplace may also lead to situations where a business’ confidential information is accidentally disclosed publicly, most likely by the inclusion of this information in prompts subsequently used to train algorithmic models (and thereby becoming part of the data used to train AI models in the future). The accidental publication of confidential information can potentially lead to substantial damage to businesses, including damage to their reputations, profits, and regulatory issues. Lawyers should therefore take care to understand what information may be inserted into algorithmic models, whether this information is confidential, and what the potential implications of the use of this information might be.
Discrimination
Employers can use AI in a variety of ways in the workplace, as outlined above – this can range from recruitment practices to performance management and dismissal. However, an algorithm is only as good as the information that is inserted into it, and if there is bias or error in what information is inserted then this can lead to potential disadvantage to particular categories of employees. An example of a way in which the use of AI might be discriminatory was suggested in a July 2023 report produced by the Ada Lovelace Institute: if an algorithm automates shift allocation based on the previous availability and productivity of workers it could potentially lead to a situation where a disabled employee, who may periodically have periods of absence from the workplace, is offered fewer shifts and therefore may earn reduced pay.
Employment lawyers will therefore need to understand how a particular employer is using AI in the workplace in order to understand what information is being used, how the AI is impacting on the workplace, and what the effect on employees and third parties might be. In addition to the example above, there could be impacts on recruitment processes, redundancy processes (and redundancy payout), the drafting and use of a settlement agreement, performance management, and how disciplinary processes are undertaken.
Constructive dismissal
The use of AI in the workplace can also conceivably lead to situations where employees’ contracts are breached by the use of, and outcomes relating to, AI applications. This could lead to resignations and potential constructive dismissal claims.
Fail to prepare, prepare to fail
Businesses should think critically about how they intend to use AI in the workplace, what information they intend to use, and what the practical effects of the use of AI might be on their employees, information, and revenues. There are substantial benefits to business in using AI (notably reducing costs, increasing efficiencies, and increasing productivity), but there are also potential legal, practical and financial drawbacks that must be carefully considered.
