I am currently in law school and will be completing my degree within the next year. With AI increasingly being used for legal research, drafting, due diligence and several other aspects of legal work, I am wondering what skills I should focus on developing before I graduate.
I don’t want to spend my remaining time learning skills that may become less valuable as technology changes. I would instead like to develop practical skills that can genuinely make me more useful to a law firm, company or other legal employer and help me stand out when I start looking for jobs.
What skills would you recommend law students learn in the current AI-driven legal industry, and where can I learn or develop these skills effectively before graduating?
If you are graduating in the next one year, I would focus less on learning a particular AI tool and more on developing your legal instinct. Tools will keep changing, but the ability to look at a set of facts and identify what is legally significant, what the real issue is, what information is missing and where the risk lies will remain fundamental. In fact, I think polishing your legal instinct is more important now than ever.
A few skills I would consciously work on are:
1. Learn to identify the real issue before researching it.
Don’t immediately turn to a database or AI when you receive a problem. First ask yourself: What is the actual dispute? Which fact is likely to matter? What does the client ultimately need? What is the strongest argument on the other side? This ability to frame the problem correctly becomes particularly valuable when AI can generate endless possible answers to a poorly framed question.
2. Learn to interrogate AI, rather than simply prompt it.
“Draft an opinion on X” is not really an AI skill. Instead, use AI to break down a problem, test competing propositions, identify counterarguments, challenge your assumptions and tell you what you may have missed. Then independently verify the output. You should be directing the tool, not allowing the tool to direct your legal analysis.
3. Become very good at verification.
This is perhaps one of the most important skills in AI-assisted legal work. Learn to trace every proposition back to the original judgment, statute, regulation or contractual provision. Don’t rely on an authority merely because AI has cited it. Check the actual judgment, the relevant paragraphs and whether the position has subsequently been distinguished or overturned. There is a significant difference between an answer that sounds legally plausible and one that is legally defensible.
4. Learn to work with large factual records.
Practise preparing chronologies, issue-wise document matrices, extracting admissions, identifying contradictions and mapping documents to particular issues. AI can help you process a large volume of material; your skill lies in deciding what actually matters.
5. Develop “document-to-argument” thinking.
Don’t stop at summarising a document. Train yourself to move from document → fact → legal issue → proposition of law → application → argument → relief. For example, if an email contains an admission, don’t just highlight it. Ask what the admission proves, which issue it affects, whether it contradicts another document and what legal consequence follows.
6. Use AI to improve your drafting, not replace your drafting ability.
You can use AI to create a first chronology, structure a research note, identify repetition or suggest alternative formulations. But you should be able to recognise when an argument is poorly sequenced, when a qualification is missing or when a sentence sounds impressive but says very little. AI can give you words; you need to develop the judgment to decide which words should remain.
7. Develop adversarial thinking.
Whenever you formulate an argument, try to formulate the best response to it. Ask yourself, “If I were appearing for the other side, how would I attack this?” You can also use AI for this purpose. Having your argument challenged is often more useful than having it reinforced.
8. Build some technological and data fluency.
You don’t need to become a programmer. But being comfortable with Excel, document comparison, document-management systems, e-discovery tools, AI-assisted research and basic data analysis will make you considerably more useful, particularly in disputes, due diligence and transactional work.
9. Develop commercial judgment.
Don’t stop at asking, “What is the legal position?” Start asking, “Given the legal position, what should the client do?” If you are reviewing a termination clause, understand its commercial consequences. If you are working on a dispute, understand the economics of pursuing it. If you are doing due diligence, understand which findings are actually material to the transaction.
Finally, don’t underestimate communication, judgment and the ability to take ownership of a problem. A client or senior lawyer will not value you simply because you can retrieve a case in seconds. They will value you because you can understand the problem, ask the right questions, identify what matters, communicate it clearly and suggest a way forward.
If I were in your position, I would therefore spend less time collecting certificates in individual AI tools and more time working through real legal problems. Take a case, identify the issues yourself, use AI to test your analysis, verify every source, prepare your argument, ask AI to attack it, refine it and then try explaining the conclusion in two minutes to someone who knows nothing about the case.
That process will build a skill set that is much more durable than simply knowing how to use any one AI platform.
