Employers facing significant rise in time and financial burdens
Disgruntled employees in Scotland are turning to artificial intelligence to make claims which are threatening to overwhelm an employment tribunals system already under acute pressure.
How AI-generated claims are increasing the burden on employers
One Glasgow-based employer, who employs around ten people in the construction industry, recently faced multiple claims of discrimination raised by one employee.
The employer faced hundreds of pages of AI-generated case law and jurisprudence on a claim for sex discrimination. When this failed to gain traction, another raft of AI documentation was filed under alleging racial discrimination.
This was rejected by an interim assessor, so the employee brought a claim on the grounds of disability discrimination, using extensive AI-generated case law.
“We try to be a fair employer and had many years of excellent relationships with our workers. It took up a huge amount of time in a small business. It seemed to me that they were trying one complaint, and when that was knocked down, they went on the offensive with another one, all using AI engines with some obscure employment law references. It’s been a nightmare,” said the managing director, who did not want to be named.
The employer spent many hours and multiple thousands of pounds on legal fees, while the employee was representing themselves at no extra financial cost to them.
What is ‘agentic flooding’?
The Economist magazine recently reported that “AI-induced demand is overwhelming bureaucracies built for the analogue age”, and this includes all kinds of courts and tribunals. The inundation of AI-generated complaints is now being called ‘agentic flooding’.
“Free, AI-powered legal advice should be good news for workers. Instead, it is proving to be a tragedy for the commons. For workers with genuine grievances, the surge in demand means long waits for justice,” said the magazine.
Employment tribunals in Scotland are more susceptible to AI claims than other parts of the justice system because they are open to lay persons to make claims. The system was deliberately set up in the 1960s to allow access and resolve disputes more quickly and cheaply.
The problem is not the use of AI itself, but its consequence. AI tends to make even the most straightforward workplace disputes excessively legalistic and length
— Fiona Cameron
Why AI is making workplace disputes harder to resolve
Fiona Cameron, partner in employment law at Gillespie Macandrew, says: “AI-generated complaints and grievances are here and are real. It’s a phenomenon that is spilling into the tribunal system too, with claimants frequently using AI to frame tribunal claims and correspondence.
“The problem is not the use of AI itself, but its consequence. AI tends to make even the most straightforward workplace disputes excessively legalistic and lengthy. Employers must sift through the now recognised term ‘AI slop’ to identify the substance of a complaint before investigating and resolving it,” she adds.
More time is spent reviewing documents in order to tease out the relevant points. What might previously have been three or four bullet points is now a 30-page letter
— Ewan Stafford
Where do you want your philanthropy to be active?
Key set-up decisions depend on whether you want to operate in Scotland, the rest of the UK, overseas, or a mixture. You want your trustees to be able to focus on achieving your philanthropic goals – not caught in a regulatory quagmire.
The time and cost implications for Scottish employers
Ewan Stafford, an employment law partner with Harper Macleod, says the legal consensus seems to be that with AI being so prevalent the human element has been lost.
“Employees seem more likely to take grievances and complaints through formal routes rather than seek early, informal, non-process-driven resolution,” he says.
He says this is particularly true of grievances where it seems easier to feed some broad assertions into an AI engine and get multi-paged documents which repeat legal points in various different ways.
“What this means on a practical level is that much more time is spent reviewing documents in order to tease out the relevant points. What might previously have been three or four bullet points is now a 30-page letter,” he explains.
He says AI-generated documents frequently contain legal terminology and adopt adversarial language that may exaggerate issues in workplace disputes.
“Minor issues can become lengthy formal grievances that are considerably more difficult to investigate and resolve. This has cost implications for the employer.”
This has implications for the employee too who wants to raise a serious issue but has been blinded by AI’s legal outputs.
“With a reliance on AI, an employee does not seem to know what their position actually is. ChatGPT may seem articulate on the surface, but when it comes to the employee explaining what it is they are looking for and why, either at a hearing or in a meeting with their employer, their position quickly falls apart.
This is resulting in time, money and energy being wasted for the employers and their advisers having to sift through the 30-plus pages to understand and untangle them.
How should employers respond to AI-generated grievances?
Cameron says: “Our advice is that businesses concentrate on substance, not volume. Employers should resist the temptation to respond to every single allegation, which only makes a complaint or claim feel even more unwieldy and can further entrench positions. Instead, employers should identify the issues genuinely requiring investigation and address them proportionately.”
Stafford’s suggestion is that those genuinely concerned with their employment situation should use a more concise approach.
“Quality is always better than quantity but that is an art that AI is yet to master. Where employees are using AI to craft documents for them why wouldn’t they submit tighter documents that they think support their position?
“I would suggest that the AI is still very much in its infancy but the use of it to bombard clients with papers is something which is readily recognised. It means that greater time and cost is incurred in picking apart the submissions to identify what the key issues are.”
However, he warns Scotland’s employers are still bound by existing legislation and need to deal with grievances and complaints appropriately.
“The use of AI is not limited to the workplace as it is also something that we are seeing more of in employment tribunals. Claims are more complex and often introduce complex legal elements, such as discrimination or whistleblowing, which when the papers are reviewed have little prospect of success.”
He suggests more judicial resources must be applied to the case management of claims in order to identify the issues.
How employment tribunals are responding to AI use
The employment tribunals recognise the issue of ‘agentic flooding’ and are taking steps to manage AI use. Tribunals in Scotland and England and Wales recently issued joint guidance following a significant rise in interim-relief applications, many apparently involving AI-generated material.
Existing safeguards include active case management and powers to strike out vexatious claims or those with no reasonable prospect of success.
A Ministry of Justice spokesperson said: “We’re aware of the growing use of AI in employment tribunal cases and we’re taking steps to ensure any use of the technology in courts and tribunals is safe, appropriate, and properly controlled.
“That includes boosting the support given to employment tribunals through additional judges, legal officers, modern case management systems, and remote hearing technology.”
The Scottish Courts and Tribunals Service was approached for comment.
AI is here to stay; resisting it is not the answer. Businesses, lawyers and tribunals must work together to ensure legitimate claims are heard efficiently and avoid consuming resources that they simply do not have to spare.