AI IN LAW
Law runs on language, precedent and knowing where the edge cases hide. That makes legal work an obvious target for large language models, and just as obviously dangerous when a fluent sentence masks a missing clause or a fabricated citation.
Inside commercial firms and in-house teams, the real shift is less about replacing lawyers and more about reorganising the bottom rungs of the ladder. Tasks like first-pass contract review, disclosure sorting and summarising long bundles used to be how trainees paid their dues and learned what a bad clause looks like in the wild.
If you are studying law now, simply knowing how to prompt a tool is no edge at all. Your value comes from verification, commercial context and understanding how a firm turns its own institutional knowledge into trusted work. The students who stand out treat AI output like opposing counsel’s draft: useful to read, never taken on trust.
Questions every law student should be asking
- 01
How do you verify every authority, defined term and cross-reference in an AI-assisted draft without skimming past plausible errors?
- 02
If first-pass document review shrinks for trainees, where will junior lawyers get the repetitions needed to spot commercial risk instinctively?
- 03
Where does confidential client matter data go when pasted or synced into a legal AI platform, and what do engagement terms allow?
- 04
When firms bill by the hour versus fixed fees, how do incentives shift once research and drafting take a fraction of the old time?
- 05
Is the real competitive advantage the model vendor, or the way a firm organises and guards its own precedent bank?
Want a repeatable checklist for testing any vendor claim or headline in law? Read our full guide on how to think critically about AI →
Articles on Law
- AI in law: junior jobs are fine, the training isn't
City trainee numbers flat for five years, despite legal AI
- Legal AI tools for law students, and how to check them
About 1,600 court filings now logged with AI hallucinations
- AI in law: courts are punishing fake AI citations
AI hallucination cases in court: 120 in May, 712 by December
Key terms for Law
A large language model (LLM) is an AI system trained on text to predict tokens in sequence, letting it draft, summarise, translate and write code.
A context window is the maximum amount of text, code or files an AI model can hold and process at one time across both your input and its response.
An AI hallucination happens when a model generates fluent, confident output that is factually wrong, unsupported by sources or completely invented.
Retrieval-augmented generation (RAG) fetches relevant passages from trusted documents before a model answers, grounding replies in real reference text.
AI governance is the set of internal policies, audit trails, roles and approval checks an organisation uses to control how AI tools run in practice.
The EU AI Act is European legislation regulating artificial intelligence by risk tier, banning certain practices and setting duties on high-risk uses.
Natural language processing (NLP) is the branch of AI enabling computers to parse, search, classify, translate and generate written or spoken language.
Human in the loop is a workflow design where a qualified person reviews, edits or approves AI outputs at high-stakes checkpoints before action runs.
Guides
Practical playbooks for law students preparing for coursework, internships and graduate roles alongside AI.
A practical five-step framework for students and juniors to evaluate any AI headline, vendor demo or workplace tool without falling for hype.
How to use AI at university within academic rules: checking department policies, disclosing assistance properly and learning instead of outsourcing.
The practical AI skills graduate employers ask about in interviews, from problem framing and verification to workflow design, and how to prove them.
A repeatable step-by-step method for students and juniors to verify AI outputs, catch fake citations, check numbers and audit confident mistakes.
Get future law articles by email
Written by students, every two weeks. Grounded in primary sources and real workflows.