Software Development · Legal

Security threat model for Legal

Use this before building or launching a feature that handles sensitive data. Written for a mid-sized law firm.

Your prompt is ready 2 brackets to fill
You are an expert in this kind of work, and you know the Legal sector from the inside.

## Your task
Act as an application-security engineer. Build a STRIDE threat model for [SYSTEM OR FEATURE] at a mid-sized law firm, which handles data for clients, partners and associates. List assets, entry points and trust boundaries, then each threat with likelihood, impact and a mitigation. Prioritise against client confidentiality, professional-conduct rules and jurisdiction (the output is a draft for a qualified lawyer to review, not legal advice).

## About my situation
- How we measure success: matter turnaround and billable utilisation

## What matters in Legal
- Rules and standards that usually apply: client confidentiality and privilege; bar council and professional-conduct rules; the law of the relevant jurisdiction; limitation periods and court deadlines; conflict checks; every output is a draft for a qualified lawyer, not legal advice.
- Common risks to plan for: missed deadlines; conflicts of interest; confidentiality breaches; citing wrong or made-up case law; fixed fees that run over.
- Tools commonly used: document and case management; research databases (Westlaw, SCC Online, Manupatra); Word with tracked changes.
- Words people in this field use: matter; brief; pleading; affidavit; clause; indemnity; jurisdiction; precedent; retainer.
- Rules differ by country. If I have not said where I work, ask me, and do not present one country's law as if it applied everywhere.

## What I will give you
Architecture description: [DESCRIBE ARCHITECTURE]

## How to work through it
1. Describe the system: assets, users, entry points and trust boundaries.
2. Go through STRIDE (spoofing, tampering, repudiation, information disclosure, denial of service, elevation of privilege) for each entry point.
3. Rate each threat by likelihood and impact.
4. Give a practical mitigation for each.
5. Order the mitigations by priority.

## What to deliver
- Assets and trust boundaries
- Threat table with likelihood, impact and mitigation
- Prioritised action list
- A data flow diagram in Mermaid syntax

## Rules
- Make every point specific to my situation and my industry; replace generic advice with detail that fits.
- Use plain language, short sentences, and tables where they make the answer easier to scan.
- Use only the facts I give you. Do not invent figures, laws, sources or quotes. Mark anything you had to assume with "Assumption:".
- If something important is missing, ask me up to 3 short questions before you start.

## Before you answer
Check your draft against this list and fix anything that fails:
- The code works as written: no invented libraries, functions or APIs, and versions are named where they matter.
- It is written for a mid-sized law firm; nothing in it would read the same for any other business.
- Nothing in it breaks client confidentiality, professional-conduct rules and jurisdiction (the output is a draft for a qualified lawyer to review, not legal advice).
- Every item under "What to deliver" is included.
Open in ChatGPT Claude Perplexity

How to use this prompt

  1. Copy the prompt. Press Copy prompt and the full text goes to your clipboard.
  2. Fill the brackets. Use the form above, or replace [SYSTEM OR FEATURE], [DESCRIBE ARCHITECTURE] with your own details.
  3. Paste and send. Paste it into ChatGPT and Claude and send. Ask follow-up questions to refine the answer.

Why use this Security threat model prompt?

This security threat model prompt is written for a mid-sized law firm. It tells the assistant who the answer is for (clients, partners and associates) and the limits it must respect: client confidentiality, professional-conduct rules and jurisdiction (the output is a draft for a qualified lawyer to review, not legal advice). That context is what turns a generic reply into one a Legal team can use straight away.

Most people type a single line and get a generic answer back. This prompt gives the assistant the role, the context and the structure it needs, so the first answer is already close to what you would hand to your team. Because it is written for Legal, it keeps the focus on what that industry measures: matter turnaround and billable utilisation.

How to get the best answer

  1. Be specific in the brackets: real names, numbers and dates give a far better answer than placeholders.
  2. Paste real material: where the prompt asks for notes, data or text, paste the actual content rather than a summary.
  3. Refine in the same chat: ask "make it shorter", "turn it into a table" or "add more detail on step 2".
  4. Check the result: review it against what you know about matter turnaround and billable utilisation before you share it.

Other industries

Same prompt for other industries

The same security threat model prompt, rewritten for each industry.

FAQ

Frequently asked questions

Which AI assistant works best with this Security threat model prompt?

It works in any chat assistant. It suits ChatGPT and Claude best. Paste it in, send it, and ask follow-up questions in the same chat.

What do I need to replace before sending it?

Replace the 2 highlighted brackets – [SYSTEM OR FEATURE], [DESCRIBE ARCHITECTURE] – with your own details. The more specific you are, the better the answer.

Is this prompt only for Legal?

This version is written for a mid-sized law firm. The same prompt is also available for 15 other industries, listed on this page.

Is it free to use?

Yes. Copy it as often as you like and use it in your own work. No sign-up is needed.

Can I trust the answer?

Treat it as a strong first draft. Check facts and figures yourself, especially anything with legal, financial, medical or safety consequences.

Keep exploring

Browse more Legal prompts