AI is changing how workers bring disputes to Britain's employment courts. The immediate effect is not a quieter system or simpler paperwork, but a much heavier load of claims, many of them drafted with tools such as ChatGPT or Grok.
A memo from tribunal presidents Barry Clarke and Susan Walker describes a court system under pressure. Claims rose 39 percent in the year through March 2026, and the backlog jumped 55 percent to 64,000 unresolved cases.
Free drafting changes who can file
The core shift is straightforward: workers can now produce legal claims without paying lawyers. ChatGPT or Grok can turn a workplace complaint into formal-sounding legal text at no cost to the person filing it.
That does not mean the filings are always useful. According to the source article, many AI-generated filings run hundreds of pages. They can also include fabricated laws and unrealistic demands, which creates more work for the people who must read, assess and answer them.
The result is a new pressure point for Britain's employment courts. When a tool makes filing easier, more people can enter the system. But if the filings are long, confused or based on false legal material, the system may have to spend more time separating real issues from unusable claims.
The backlog is the central problem
The numbers in the memo show why this matters. Claims rose 39 percent in the year through March 2026. Over the same period, the backlog jumped 55 percent to 64,000 unresolved cases.
Interim relief applications have surged a hundredfold. That matters because these applications add to the immediate workload of employment courts, especially when they arrive in large numbers and require attention before the underlying dispute can be fully resolved.
The cost is not only administrative. Workers with real grievances wait longer for justice. Employers also pay more to respond, whether a claim is legitimate or made up.
This is the hard part of AI-assisted access to legal systems. Lowering the cost of drafting can help people speak up. But when many filings are inflated, inaccurate or unrealistic, the burden shifts onto courts, employers and other workers waiting in the same queue.
Why AI-generated lawsuits create a shared burden
The Economist calls the situation a "tragedy of the commons, AI edition." The phrase fits the pattern described in the source article: a tool that is useful to an individual filer can create costs for everyone else using the same legal system.
For a worker, using AI to draft a claim may feel rational. It is free, fast and available without a lawyer. For the court, however, hundreds of pages of AI-generated material can become a screening problem, especially when fabricated laws are mixed into the filing.
For employers, the distinction between a legitimate claim and a made-up one may still require time and money to establish. Even weak claims can demand a response. Even unrealistic demands can create process costs.
That is why the burden spreads beyond the person using the AI tool. The court must process the paperwork. Employers must respond. Other claimants must wait behind a growing backlog.
New claim grounds could intensify the strain
The source article says the flood could get worse because Labour's new Employment Rights Act adds about 25 new grounds for claims and removes compensation caps.
That change matters in the context of AI-assisted filing. If there are more possible grounds for claims, and workers can draft claims for free, the number of filings may keep increasing. The source does not say how many additional cases will result, but it does point to the risk of more pressure on an already strained system.
The issue is not simply whether AI belongs in legal work. The problem is what happens when AI-generated documents enter courts at scale without reliable limits on length, accuracy or realism.
The same tools can help or harm courts
The source article also points beyond Britain. The US faces a similar crisis, and one federal judge called AI-generated lawsuits an "existential threat to the federal courts." That warning reflects the same basic concern: courts can be overwhelmed by filings that look legal but contain serious flaws.
But AI is not only a source of bad filings. A study from Pakistan shows a different outcome: judges with AI tools and training processed more cases, faster.
Together, those examples show that the impact of AI depends on where and how it is used. When untrained users generate long legal claims filled with fabricated laws, courts can slow down. When judges have AI tools and training, case processing can improve.
Britain's employment courts are now facing the difficult version of that transition. AI has made it easier to file, but easier filing does not automatically produce clearer claims, faster justice or lower costs. For now, the memo from Barry Clarke and Susan Walker points to a system absorbing more claims, a larger backlog and a new layer of AI-created work.