AI in practice

What law students should learn about AI before they graduate

Seven skills, starting with how the model works.

[date] · [author] · 6 min read

You will work in a profession where everyone uses AI, your client before your colleague. The difference between a lawyer these tools serve and one they trip up is not knowing which buttons to press. It is seven skills you can learn while you are still at university.

1. Know how the model writes

A language model learns from a vast amount of text to predict the next word, then is tuned toward the answers people prefer. It does not search a database or remember pages; it stores patterns. What it has seen often, it masters; what is rare, it guesses. And its knowledge stops on the date its training ends. These four facts explain most of its mistakes, and why it is fluent before it is accurate.

2. Ask a complete legal question

The governing law, the date of the facts, the facts themselves, and what you want: a text, a ruling, or steps. An incomplete question invites the model to fill the gap with the most likely answer.

3. Always go to the source

Read the whole article, and read the ruling, not its summary. What is written about a text is not the text.

4. Read the citation

Learn the format of a Court of Cassation citation: the appeal number, the judicial year, the hearing date and the Technical Office collection. If you can read a citation, you can spot an invented one in seconds.

5. Check the date

Laws are amended and repealed. The Labour Law you studied in your first year may not be the law in force when you graduate: Law 14 of 2025 replaced Law 12 of 2003 from 1 September 2025.

6. Protect client secrets

Do not put the facts of a real case into a general tool when you do not know where it keeps them or whether it trains on them. It is a professional duty before it is a technical one, and the data protection law makes it a legal duty too.

7. Write for yourself

The tool writes a draft. Arabic legal style, precise words and well ordered argument, is only learned by writing. In exams and research papers, follow your faculty's rules on AI.

Why a general model gets this wrong

The first skill explains the other six. The model gets Egyptian law wrong because Egyptian texts and rulings are a small part of what it learned from, so it fills the gap with the most common patterns from other systems. It drops qualifications because short, firm answers are usually preferred. It invents numbers because numbers are the hardest part of predicting the next word. It misses what is new because its knowledge stops on a date. A student who understands this knows when to trust the tool, and when to open the book.

In Bayan

Bayan Academy offers short free courses in Arabic that start from these skills, including "Legal research with AI: the basics" and "From answer to source", with a certificate after each course [Sherif confirms the course list]. Universities can partner with Bayan, and law and engineering students can apply to the Bayan Ambassador Program.

Share

The month's key legal updates and what's new in Bayan, once a month.

Ask Bayan your first question.

Or book a team demo and bring a real contract: thirty minutes is enough to know whether Bayan fits your practice.

Book a team demo