AI Is Clogging the Courts — and It's a Tragedy of the Commons

The Economist just published a piece titled "The tragedy of the commons, AI edition" that hit 42 points on Hacker News with active discussion. The core story: Britain's employment courts are being overwhelmed by AI-generated legal complaints.

This is a fascinating — and alarming — case study in how AI creates systemic problems when deployed at scale without governance.

What's Happening

British employment law has a provision called "interim relief" — an emergency measure where a judge can order a firm to reinstate a fired employee or pay their wages. Historically, tribunals received about 20 applications per year. These were rare, serious cases — typically whistleblowers or trade union officials.

Now, with AI tools that can generate legal complaints in seconds, the number of applications has exploded. People are using AI to mass-file employment complaints, many of them boilerplate or borderline frivolous. The system designed to protect vulnerable workers is being drowned in noise.

The Tragedy of the Commons Framework

This is a textbook tragedy of the commons:

• The commons: The employment tribunal system — a shared resource with finite capacity
• The rational individual incentive: File a complaint (AI makes it free and easy), potentially get a payout
• The collective outcome: The system becomes so clogged that legitimate cases can't get heard

Each individual filing a complaint is acting rationally — the cost to them is near-zero with AI tools. But the aggregate effect is that the system becomes useless for everyone, including the people it was designed to protect.

Why This Matters Beyond UK Employment Law

This pattern will repeat across every system that has:

1. A process for filing complaints/appeals/requests
2. No meaningful cost barrier to filing
3. AI tools that can generate plausible-looking filings

We're already seeing it in:

• Parking ticket appeals — AI-generated appeals flooding municipal systems
• Planning objections — AI mass-generating objections to construction projects
• DMCA takedowns — Already a problem, now supercharged by AI
• Tax disputes — AI generating complex tax challenges

As one HN commenter noted: "From parking tickets to planning, the state is under siege by AI."

The Lawyer Problem

The Economist piece also raises a subtler issue: if AI can generate legal filings, do we still need lawyers? The answer from the HN discussion was nuanced:

"A lawyer isn't going to sign on to take liability risk unless they spend quite a lot of time analyzing the AI's outputs for possible blunders. So it structurally can't cost significantly less."

This is key. The value of a lawyer isn't just writing the complaint — it's taking legal responsibility for its accuracy. An AI can generate a complaint, but it can't be held liable if it's wrong. So either:

1. A lawyer still has to review every AI-generated filing (cost savings are minimal), or
2. People file without lawyers (quality drops, courts get clogged)

We're seeing option 2 play out in real time.

Potential Solutions

The HN discussion surfaced several approaches:

1. Filing Fees with Compassionate Waivers

"There needs to be a financial mechanism to deter nuisance complaints. A small penalty for lodging a complaint and losing. This fee should help pay for the courts and reimburse an employer for time spent defending itself. (With the judge/magistrate able to waive the fee on compassionate grounds.)"

This is the classic Pigouvian tax approach — make the cost of filing reflect the cost it imposes on the system.

2. AI-Arbitrated Filtering

AI created the problem, but it could also help solve it. An AI system could pre-screen complaints and flag:

• Boilerplate filings (likely AI-generated)
• Cases with no legal merit
• Patterns of mass-filing from the same source

3. Rate Limiting

Just as APIs rate-limit requests, courts could rate-limit filings per person per time period. This would prevent mass-filing while preserving access for legitimate cases.

4. Verification Requirements

Require filers to verify key facts under penalty of perjury, with AI-generated content flagged for additional scrutiny.

The Broader Pattern: AI as a Commons Disruptor

What's happening in UK employment courts is a preview of a much larger problem. AI dramatically lowers the cost of generating content — legal filings, support tickets, product reviews, academic papers, news articles. But it doesn't lower the cost of processing that content.

Every system that receives user-generated content — courts, customer service, academic journals, review platforms — is about to face the same crisis. The cost to create content approaches zero. The cost to review it stays constant. The system breaks.

This isn't just about AI being annoying. It's about the fundamental economics of information systems. When creation costs drop but review costs don't, the system either:

1. Raises barriers to entry (filing fees, verification, rate limits)
2. Automates review (AI judges AI — with all the risks that entails)
3. Degrades (the system becomes useless under the volume)

UK employment courts are showing us option 3 in real time. The question is whether other systems will learn from this before it's too late.

---

This is another example of AI creating systemic risks that require governance, not just technology, to solve. What other systems do you think are vulnerable to AI-induced tragedy of the commons?