You’ve been wronged by a company. You know it. They know it. And yet nothing happens — because who has $300 an hour to chase $4,000?
That asymmetry — the one where corporations can afford lawyers and you can’t — has been the quiet engine of injustice for as long as anyone can remember. Not because the law is too complex for ordinary people. But because the gatekeepers who translate it charge more than the claim is worth.
So when a Norwegian airline delayed my luggage into oblivion across an ocean and two jurisdictions, I did what any rational person would do: I Googled “how to sue an airline,” felt my soul leave my body at the sight of international aviation law, and almost gave up.
Then I opened ChatGPT.
The tool that everyone said would replace lawyers actually works best when paired with someone stubborn enough to double-check its work. That’s the paradox nobody’s talking about: AI doesn’t replace legal judgment — it replaces the paywall around it.
Here’s what happened. I described the situation in plain English — flight details, what happened to the bag, what the airline said (nothing), what EU regulation 261/2004 covers, what the Montreal Convention says about international baggage. Within minutes, ChatGPT had mapped the entire procedural landscape: which Norwegian authority handles consumer complaints, what forms I needed, what language to file in, what deadlines applied, and what specific legal language to use when the airline inevitably stonewalled.
Things that would have taken a lawyer three billable hours to explain took ChatGPT ninety seconds. And it didn’t just summarize — it walked me through the sequence. Step one: file with the Norwegian Consumer Council. Step two: if no response within two weeks, escalate to the Civil Aviation Authority. Step three: small claims filing in the Norwegian system as a non-resident.
Was I nervous? Absolutely. Every headline I’d read about AI in law was about some attorney who cited hallucinated case law and got sanctioned by a judge. But here’s the thing none of those headlines mentioned: in small-claims and consumer disputes, the bar isn’t whether your legal reasoning is airtight enough to survive appellate review. The bar is whether you can navigate the paperwork at all.
Most people don’t lose small-claims cases because their argument is weak. They lose because they never file — the system is designed to exhaust you before you begin.
The airline, predictably, ignored my first two filings. This is standard corporate strategy in low-stakes disputes: ignore, delay, hope the claimant gives up. What they didn’t count on was that I had an AI assistant that could draft escalation letters in Norwegian, cite the specific regulatory clause they were violating, and generate a formal small-claims complaint formatted to Norwegian court standards — all for free.
When the judgment came — $4,760, including compensation for the lost luggage, consequential damages, and interest — I sat in my apartment in New York and stared at the screen. I had just won a legal dispute in a foreign country’s court system without hiring a single lawyer.
But let me be clear about what actually happened here, because the headline oversells it.
ChatGPT didn’t win the case. I won the case — by verifying every citation it gave me, by cross-checking every procedural step against the actual Norwegian government websites, by knowing when to trust the AI and when to override it. There were at least three moments where it confidently told me something that was technically plausible but procedurally wrong — like insisting I could file electronically when the specific court required physical mail. I caught those because I treated it like a brilliant but overconfident junior associate, not like an oracle.
The breakthrough isn’t that AI can practice law. It’s that AI can flatten the knowledge gap between you and the corporation that wronged you — and in low-stakes disputes, that gap was the only thing standing between you and justice.
The debate about AI in legal contexts always swings between two extremes. On one side: “it’ll replace all lawyers.” On the other: “it hallucinates case law and will get you sanctioned.” Both miss the vast middle ground where most real legal problems actually live — the consumer disputes, the small-claims filings, the bureaucratic nightmares that are too small for a $400/hour attorney and too complex for a layperson to navigate alone.
That middle ground is where millions of people get screwed every year. Not because they’re wrong. Because they can’t afford to be right.
ChatGPT didn’t give me legal expertise. It gave me something more valuable: the map. The territory was still mine to walk.
Justice was never supposed to be a luxury good. AI won’t fix the system — but it might just make the system’s gatekeepers irrelevant for the cases that matter most to ordinary people.
FAQ
Q: Didn't you worry about ChatGPT hallucinating case law and tanking your case?
A: Yes, constantly. But I treated it like an overconfident junior associate, not an oracle. Every citation got cross-checked against actual Norwegian government sources. It made at least three procedural errors I had to catch — like claiming I could file electronically when the court required physical mail. The AI gave me the map; I walked the territory.
Q: Does this mean ordinary people can now handle legal disputes without lawyers?
A: For low-stakes consumer disputes — delayed flights, lost luggage, small-claims filings — yes, probably. For anything with real exposure or complex liability, no. The breakthrough is specifically in the niche where traditional legal help is too expensive to justify but the procedural barriers are too high for a layperson to clear alone.
Q: Isn't this just going to flood courts with AI-generated slop lawsuits?
A: That's the fear, and it's not baseless. But the counterpoint: corporations already flood the system with resources designed to exhaust ordinary claimants into giving up. If AI tips that balance even slightly toward the little guy, that's not slop — that's access.