The Problem
AI will fill in gaps with "reasonable assumptions" when information is missing. For legal work, these assumptions can be wrong and dangerous. You might get an answer based on the law of another country when you needed the law of England and Wales, or advice for a public company when your client is private.
The Solution
Add this explicit instruction to your prompts:
"Don't make any assumptions. If anything is unclear, ask me first."
This simple addition changes the AI's behaviour from "helpful guesser" to "careful clarifier."
Implementation
Add to the beginning or end of your prompt:
[Your prompt here]
Important: Don't make any assumptions. If any details are unclear or
missing, ask me for clarification before providing your answer.
Before You Rely on This
The examples on this page are made up to show the technique. No test results have been published for it, so try it on your own work and judge it by what you find.
This guide is about how to word a prompt. It does not make a client's document or information safe to put into an AI tool. What you may enter depends on the plan your firm has, its agreement with the vendor and your firm's policy: Claude Projects for Client Matters sets out the checks for one tool. To practise, use a precedent or a published document that contains no client information, and refer to parties by role.
Why It Works
LLMs are trained on reward models that prioritise being helpful. When faced with ambiguity, they default to providing some answer rather than none. This explicit instruction overrides that training and gives the model permission to ask rather than guess.
Example
Without the instruction:
Review this employment contract for red flags.
[attaches contract]
AI assumes: the law of England and Wales, standard employment relationship, private company, no special circumstances.
Actual situation: Zero-hours contract, gig economy worker, requires IR35 analysis.
With the instruction:
Review this employment contract for red flags.
[attaches contract]
Don't make any assumptions. If anything is unclear, ask me first.
AI responds:
- What jurisdiction applies?
- Is this for an employee, worker, or contractor?
- Are there any specific concerns (IR35, restrictive covenants, etc.)?
- What industry/sector is this for?
You provide the details, and the analysis is of the contract you have, not the one the AI assumed.
Pro Tips
- Combine with "The Clarification Closer" for maximum effect
- Essential for cross-border work where jurisdiction matters
- Particularly important for technical areas (tax, IP, regulatory)
- Use in ChatGPT Custom Instructions to make it default behaviour
Common Mistakes
❌ Being too vague yourself: "Review this contract. Don't assume anything." (You haven't given the AI any context to work with)
❌ Contradicting yourself: "Assume the law of England and Wales applies. Don't make assumptions." (Mixed signals)
✅ Providing some context, but inviting questions on gaps: "This is an employment contract for a senior role, governed by the law of England and Wales. Don't assume anything else - ask me."
Quick Reference
Copy this template:
[Your legal task]
Context: [Brief relevant facts you know]
Don't make any assumptions beyond what I've stated.
If you need more information, ask before answering.