Lawyers fined for citing fake case law invented by AI: what happened and how to avoid it

What happened in the Canary Islands

February 2026: 48 judgments that didn't exist

By an order of 10 February 2026, the Criminal Chamber of the High Court of Justice of the Canary Islands fined a lawyer 420 euros for including in an appeal 48 non-existent case law citations suggested by an artificial intelligence tool.

The lawyer admitted the facts and accepted responsibility. Even so, the court found the conduct contrary to the duty of truthfulness and the rules of good faith, an abuse of the public justice service, and referred the case to the bar association in case disciplinary action was warranted. Its ruling left two ideas worth remembering: human oversight is the "backbone" of AI use in legal practice, and the tool must be understood "as assistive and never as decisive".

July 2026: again, and double the fine

Months later, the same Chamber fined another lawyer 840 euros for filing an appeal that cited at least 25 Supreme Court and Constitutional Court judgments that were wrong, apocryphal or simply non-existent. The citations appeared in quotation marks and highlighted, as if they were genuine, and the court's own Documentation Service had to check them one by one.

The court spoke of neglect of basic professional duties through gross negligence and recalled that using AI tools doesn't exempt anyone from checking information before filing it. As a detail, the amount of the fine was calculated from the annual cost of a legal AI platform. In this case too, the matter was referred to the bar association.

Why AI makes up judgments

A language model doesn't search a database: it generates the most likely text based on what it has learned. Ask it for case law backing an argument and it will produce references in the exact format of a real judgment (court, number, date, reporting judge) even if none exists. It isn't lying on purpose; it simply can't tell the plausible from the true.

That's why a generic AI tool is useful for drafting or summarising but dangerous as a source. When the result is going to a court, every citation has to come from a verifiable source and pass before a lawyer's eyes.

What the profession itself says

The Spanish General Council of Lawyers presented its White Paper on Artificial Intelligence and the Legal Profession in January 2026. According to figures released at its launch, 60% of professionals already use applications with artificial intelligence.

The document's central principle is human oversight: AI complements legal work rather than replacing it, and responsibility for strategy and filings remains with the lawyer. The two Canary Islands orders apply exactly that idea.

The problem isn't AI, it's using it as an oracle

A well-designed digital employee doesn't replace the lawyer's legal research: it handles the work that requires inventing nothing. It receives email, identifies the matter, gathers documents, extracts data from them, drafts on the firm's own templates and calculates with rules whatever needs calculating. And nothing goes to a client or a court without a person approving it. We explain it in detail in AI in law firms: what can be automated today and what can't.

How to use AI in your firm without ending up fined

1. Separate drafting from citing. AI can help you order arguments or polish a text; citations always come from an official source.
2. Check every reference in an official case law database before including it.
3. Use tools connected to sources, which show where each answer comes from, instead of a generic chat.
4. Define what a lawyer must always approve: filings, client communications and calculations with consequences.
5. Leave a trail. Knowing who prepared what, with which data and who approved it protects the firm.
6. Train the team. The EU AI Act already requires people using these systems to be sufficiently trained.
7. Delegate first what needs no external sources: incoming email, documents, data and calculations. That's where AI saves the most hours with the least risk.

To see how this works in practice, automation and AI for law firms explains what a digital employee does in a firm today.