The ever-increasing use of AI is one of the hottest topics around at present. A simple internet search for the term shows how widespread its impact is being felt at what is only the initial stage of its development. Stories about AI being able to manipulate elections and interfere with democracy; concerns over students being able to use online AI tools to pass university assignments; to Tom Hanks believing that his career will be able to live on beyond his death through AI recreating his image, point to 3 diverse areas where the effects of AI are being felt. Rumour has it, AI bots can also come up with some pretty decent employment law blogs too…
In regards to the world of work, the use of AI has been a more prominent issue over the past few years. The 2022 BBC documentary “Computer Says No” highlighted the potentially discriminatory impact which reliance on AI can have where recruitment decisions are taken exclusively by technology. It also reported on the decision taken by Mac Makeup to make 3 employees redundant entirely on the basis of an algorithm which assessed their facial expressions during a video interview, a decision which led to claims being raised against the business. The idea that such a decision could be made without any human input seems incredible, yet highlights how available AI technology is to become involved in day to day operation of a business.
The efficiencies which AI can offer is clearly an attraction, with the ability to save costs and, potentially, remove the risks which can be caused by human error. For employees, AI has the ability to make their lives easier by performing straightforward or repetitive tasks. Just yesterday, BT announced that as part of its intention to reduce its workforce by 55,000 employees by the end of the decade, a fifth of those posts would be replaced by AI.
It has been long recognised that employment law may not be able to keep up with the pace of development of AI. However, this week a bill has been introduced to the UK parliament which attempts to take the first legislative steps to try and afford some rights and protections for employees and workers whose working opportunities and circumstances could be impacted by the use of AI. This bill largely reflected concerns and recommendations made in a TUC report from March 2021.
The Artificial Intelligence (Regulation and Workers’ Rights) Bill was put forward by Labour MP Mick Whitley and approved at its first reading stage. This means that the bill will now be given greater scrutiny by parliament. While it is likely that it will take until next year for the bill to be passed into law (should that happen), the proposed rights the bill will introduce are ones which raise interesting issues which employers would be well advised to give some consideration to at this stage.
In summary, the bill would seek to introduce the following specific rules:
- A right to a human review of decisions made by AI;
- A right to human contact when high risk decisions are made in the workplace;
- A statutory duty for employers to consult with employees and their trade unions before introducing AI into the workplace; and
- An amendment to the Equality Act making discriminatory data processing expressly unlawful.
In presenting the bill, Mr Whitley highlighted one particular example of where AI had been applied in an unfair manner. Amazon had made use of an AI tool which would sift through applications for jobs. The algorithm which was used by this tool had learnt that the majority of hires previously made by Amazon for the job in question had been male. As a result of this, the AI tool had taught itself to downgrade applications by females. This situation is one which would already leave employers open to claims of direct discrimination on grounds of gender from any applicant who was unsuccessful on the basis of the AI tool’s determinations.
In situations where AI is used to determine which employee should be dismissed, there is also a real question as to whether the employer would be able to show that such a decision was one which fell within the range of reasonable responses when defending any subsequent unfair dismissal claim. How easy would it be for an employer to show reliance on an algorithm was a fair way to conduct a disciplinary or redundancy process?
While there has been a lot of debate about the moral and ethical dilemmas which increased use of AI can cause, it would appear certain it will play an ever more protuberant role in the workplace in coming years. Whether this bill is passed into law or not, there are already risks which employers need to be aware of when relying on AI to take employment-based decisions, as both Mac Makeup and Amazon have found to their cost.
I hope this blog has been of interest and gets you thinking. I promise it was written by me and not ChatGPT.
If you need any employment law advice, please get in touch with Blackadders’ Employment Law team working in Aberdeen, Dundee, Edinburgh, Glasgow, Perth and across Scotland.