Work · case studies
AI mistakes that cost real careers
30-second gist~30s read
There's a clear pattern in the AI-related career stumbles of 2023-25: someone trusted AI output as a final answer, didn't verify, and got found out. The damage isn't always being fired. Sometimes it's just losing trust with a colleague who'd quietly noticed the AI got it wrong.
The cure is simpler than the diagnosis: verify before you publish, cite, or forward. Five minutes saves a career.
If you want more
The classic case: Mata v. Avianca
In 2023, a New York lawyer used ChatGPT to research a brief. ChatGPT cited six previous cases that perfectly supported his argument. Every case had been fabricated by the AI — judges, courts, quotations. The lawyer filed it anyway. The judge found out, sanctioned him, and made the case a textbook example. The damage wasn't the AI. The damage was filing without verifying.
By 2025 the pattern has been repeated by lawyers, journalists, consultants, students, and corporate analysts in dozens of public cases.
The one habit that prevents almost all of this
- For citations: click each one before submitting.
- For statistics: find a second, named source for any number you'd quote.
- For names and dates: a 30-second Google.
- For anything you'd be embarrassed to be wrong about: ask the AI "what's the strongest reason this might be wrong?" and read the answer.
If you treat AI as a fluent collaborator who occasionally bluffs, you'll avoid almost all of the public failures.