Artificial Intelligence (‘AI’) is able to generate a legal notice in seconds, it can also draft a plaint, suggest contractual clauses, summarize facts and even direct you to the relevant authorities.
But there is a more difficult question: does a technically good draft automatically make a good legal draft?
That’s when legal drafting becomes more than just getting the right words on paper. A good brief requires an understanding of the law, an ability to determine what really matters in the facts, an ability to anticipate what the other side would present, an assessment of risk, and strategic choices. A lot of these tasks can be helped by AI, but it is still the lawyer’s final judgment
A legally correct draft may not be Suitable to the Facts of the case
One of the biggest misconceptions about AI assisted legal drafting is that a draft is good solely because it sounds legally correct. An AI tool could write a well-structured notice of breach of contract with the relevant clauses, allegations, demands, and legal threats. But the lawyer still has to ask what the client really wants to get done.
Or maybe the customer wants to renegotiate the contract instead of terminating the relationship. It could make that much harder where, for example, there was an aggressive notice of repudiatory breach threatening immediate proceedings. This is where the legal judgment comes in. AI may help with the language and the structure, but the lawyer has to determine whether the draft is appropriate to the facts, the client’s goal and the strategy.
Importance of right prompts and instructions
Artificial intelligence can speed up the time-consuming exercise of drafting legal documents. It can organize facts, prepare a preliminary draft, suggest clauses, improve language, summarize judgments, and flag possible issues. But the lawyer still has to decide what facts are important, what the document is supposed to do, whether a clause really protects the client, and whether a judgment supports the legal proposition.
Let’s take an example of drafting a complaint. AI is able to structure the facts, but the lawyer still has to find the cause of action, decide what needs to be pleaded, check the relief, check the Limitation Act, and identify any admission that could be damaging to the client later. The draft also needs to be checked against the actual documents and the client’s instructions. While artificial intelligence can help accelerate the drafting process, it’s the legal knowledge and practical judgment that can turn a draft into a document ready to present before the court.
The real value of experience is knowing what not to draft
Knowing what to leave out is one of the most important aspects of legal drafting. A lawyer might be tempted to include every possible argument, fact or authority, thinking that more detail will make the draft stronger. But experienced drafting often comes down to a simple question: if I put this in, could it cause a problem for the client later?
For instance, an overly broad indemnity clause in a commercial agreement can lead to ambiguity, increased liability or tougher negotiations. Similarly, a complaint that contains all of the factual allegations may be longer, but not necessarily more persuasive. Good drafting is not a matter of saying everything. It’s knowing what to say, what not to say, and what could damage the interest of the client if it’s mentioned in the draft.
Experience changes how lawyers read the same facts
You can give two lawyers the same instructions and get very different drafts. The difference is often in how they weigh the client’s goal, the risks and what could happen once the document is filed or signed. A litigator with experience might spot a damaging admission in a client’s statement, while a transactional lawyer might spot an ambiguity in a contract that could be the source of a dispute years later. A lawyer who knows a particular court might also realize that certain phrasing could lead to an objection or unnecessary complications.
These are more than differences in style of writing. They are the product of practical experience and the judgment derived from dealing with actual cases, negotiations and disputes. AI is able to find patterns from the existing legal material, but the lawyer still needs to connect the draft to the particular client, facts, forum, transaction and objective. That is where your legal knowledge and experience really still counts.
The biggest AI drafting risk: trusting a confident answer
AI-generated legal content can be convincing even if it’s wrong. This is why verification is not dispensable. Never assume an AI-generated citation, statutory provision, case summary, or legal proposition is correct. Before using the draft, the lawyer should go back to the original sources and records to check the important details in the draft.
If a pleading, notice or agreement is AI-assisted, the lawyer should check:
- Material facts to the source documents and the client’s instructions.
- Statutory provisions against current law.
- case citations against actual judgment
- Legal propositions to verify that the cited authority does support the point.
- The dates may be dependent on them due to limitation and procedural requirements.
- Names, figures and contractual references against original papers.
The fact that an AI tool helped prepare the document does not relieve the ultimate responsibility. The lawyer who reviews, approves and sends the draft is responsible for the accuracy, legal soundness and suitability for the client of the draft. AI is able to help with speed and structure but it cannot accept any responsibility for the advice or document issued in the lawyer’s name.
The future is for lawyers who know how to use AI
It is not the question that whether AI would replace lawyers. It’s knowing where artificial intelligence can help and where the judgment of the lawyer is key. A pragmatic approach to AI assisted legal drafting would be:
- Step 1: Maintain confidentiality: Do not upload confidential client information, privileged communications, personal data, case documents, or sensitive commercial information unless you are authorized to do so and the tool has appropriate privacy and security safeguards.
- Step 2: Know Your Goal: Before you get into AI, understand what the client is trying to accomplish.
- Step 3: Get your facts right: Distinguish between verified facts and assumptions, allegations and information still to be verified.
- Step 4: Find the law: Please conduct your own research on the legislation, rules and relevant authorities before relying on an AI generated draft.
- Step 5: Have AI write the first draft: Use it to organize facts, structure the document, simplify language, spot holes, or suggest alternative wordings.
- Step 6: Challenge the draft: Ask what is missing and what is superfluous and if there is anything in the draft that can be used against the client.
- Step 7: Review the details: Verify facts, provisions, authorities, dates, figures and other important details from reliable sources.
- Step 8: Rely on your gut: Develop the draft to fit the client’s goals, legal strategy, risks and practical implications.
- Step 9: Look at it from the other side: Before you sign off on the document, ask yourself what the opposing lawyer could challenge or interpret differently.
It doesn’t matter whether a document is written entirely by a lawyer or with the assistance of AI, the lawyer must be able to stand behind every single sentence that leaves the lawyer’s office in their name.
Strategic Legal Judgment: The Skill Behind Effective Drafting
Artificial intelligence can help to speed up the preliminary draft, but speed doesn’t necessarily mean a correct final draft. The real skill is knowing how to question the AI generated draft, check the law and authorities, spot what is missing, identify risks and reshape the document around the client’s real objective.
The lawyers who’ll benefit most from AI would be the ones who will know what to ask, what to check, what to reject and what to change. Whether a draft written with the help of AI is simply well written or is fit for purpose will remain a question of legal knowledge, practical experience and sound judgment.
Want to strengthen your legal drafting skills?
Explore Lawctopus Law School’s 6-Month Course on Mastering Litigation and Becoming an Independent Litigator to build practical litigation skills, or Mastering Legal Drafting for High-Paying Roles to strengthen one of the core skills every litigator needs.
Disclaimer: This article is intended for educational purposes only and does not constitute legal advice. Professional conduct, enrollment and practice requirements may change. Readers should refer to the applicable legislation, Bar Council rules and current professional requirements before acting on the information discussed above.
References
- Bar Council of India, Rules on Standards of Professional Conduct and Etiquette, Part VI, Chapter II, Standards of Professional Conduct and Etiquette, rr 15, 17, Bar Council of India Rules on Standards of Professional Conduct and Etiquette.
- Bharatiya Sakshya Adhiniyam 2023, s 132
- American Bar Association, Standing Committee on Ethics and Professional Responsibility, Formal Opinion 512: Generative Artificial Intelligence Tools (29 July 2024), ABA Formal Opinion 512
- Indian Contract Act 1872, ss 124–125
- Code of Civil Procedure 1908, sch 1, Orders VI and VII
- T Arivandandam v TV Satyapal (1977) 4 SCC 467
- Limitation Act 1963, s 3
- National Institute of Standards and Technology, Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile, NIST AI 600-1 (July 2024)
About the Author
Vishakha Girhe is a Learning Manager at Lawctopus Law School. She holds an LL.M. in Business Laws and has practised as an independent litigator at the Nagpur District Court and Bombay High Court, Nagpur Bench.
