ChatGPT Prompts for Family Law Attorneys: Privilege-Safe Client Communication for Divorce & Custody Practice
Family law is a different kind of practice. Every message you send goes to someone in the middle of the worst year of their life — losing time with their kids, splitting a house, negotiating over a retirement account they spent twenty years building. The writing has to be precise enough to survive scrutiny and human enough that a frightened, angry, or grieving client can actually absorb it. And there's a lot of it: the custody-schedule explainer, the discovery chase, the settlement offer broken down in plain English, the phone-call recap after a hearing that didn't go the way anyone hoped.
That's exactly the kind of repetitive, high-stakes writing ChatGPT can accelerate — and exactly where one careless prompt creates a problem no time savings justifies. Because here's the governing fact most "AI for lawyers" advice buries: a growing body of case law and ethics guidance treats conversations with general-purpose AI tools as NOT protected by attorney-client privilege and potentially discoverable in litigation — including divorce and custody proceedings. This article gives you the privilege-safe baseline and 12 ready-to-adapt prompts for the day-to-day practice of family law.
TL;DR
ChatGPT can draft custody-schedule explainers, discovery cover letters, settlement breakdowns, and post-hearing summaries in seconds — but only safely under three rules. First: nothing identifying ever enters the prompt — no client name, no child's name or school, no financial account numbers, no case numbers. AI chat logs are potentially discoverable; draft with [Client], [Matter], and [FIGURE] placeholders, and personalize inside your case management system. Second: AI drafts communication, never legal strategy — anything touching strategy (settlement positions in disputed matters, custody arguments, enforcement decisions) is attorney work that at most gets formatted by AI, never generated. Third: supervising-attorney review is mandatory — ABA Model Rule 5.3 treats AI like any nonlawyer assistance, and ABA Formal Opinion 512 (2024) makes the confidentiality duty explicit. Below are 12 prompts organized around post-signature practice — this is not an intake pack — plus the one-line rule for the whole office: if it would hurt to see it read back in court, it doesn't go into the chat window.
The privilege-safe baseline, before a single prompt
- AI chats are not privileged. Treat every prompt as potentially discoverable.
[Client]/[Matter]placeholders always; never a name, a child's school, an account number, or a case number. - Communication yes, strategy no. AI helps you say clearly what you've already decided. It never decides — not a settlement position, not a custody argument, not whether to file for contempt.
- Model Rule 5.3 review, every time. Drafts flagged for attorney review mean exactly that: the supervising attorney reads before anything reaches a client, opposing counsel, or a court.
Custody & parenting plan communication (3 prompts)
1. The parenting-schedule explainer
Rewrite this parenting schedule in plain, neutral language for [Client]: [PASTE THE SCHEDULE TERMS]. Structure by regular weeks, holidays, and summer. No editorializing about fairness — just what happens when, and what [Client] needs to do at each exchange. 8th-grade reading level.
2. The adverse-ruling delivery
Draft a message to [Client] delivering an unfavorable custody ruling: one clear sentence on what the court decided ([GENERIC OUTCOME]), one on what it does and doesn't change day-to-day, one on the options we'll discuss at our next call ([DATE]). Compassionate, honest, no false hope, no criticism of the court. ATTORNEY REVIEW REQUIRED before sending. Under 140 words.
3. The co-parenting communication reset
Write guidance for [Client] on keeping written communication with their co-parent child-focused and court-appropriate: three short principles (brief, factual, about logistics; never about the relationship; assume a judge may read it), each with a one-line example. Under 150 words.
Financial discovery & documentation (3 prompts)
4. The disclosure checklist cover
Write a cover message to [Client] accompanying our financial disclosure checklist: why complete disclosure matters (credibility with the court, avoiding delays and re-dos), the deadline ([DATE]), and an offer to walk through any item on a call. Firm about the deadline, warm in tone. Under 120 words.
5. The document chase
Write a follow-up to [Client], who hasn't sent the [DOCUMENT TYPE, e.g., last three bank statements] we requested on [DATE]. One line on what's still missing, one line on what the delay affects ([CONSEQUENCE, e.g., our response deadline]), and the easiest way to send it. Patient but clear. Under 90 words.
6. The QDRO plain-English explainer
Explain in plain English what a QDRO (qualified domestic relations order) is and why dividing a retirement account requires one: what it does, roughly how the process flows, and why it takes time. Generic education only — no advice about [Client]'s specific accounts, no tax guidance beyond "we'll flag tax questions for your financial professional." Under 150 words.
Court dates & case status (3 prompts)
7. The hearing prep reminder
Write a hearing-preparation message to [Client] ahead of [HEARING TYPE] on [DATE] at [TIME], [LOCATION]: when to arrive, what to bring, courtroom conduct in three short bullets, and when we'll meet beforehand ([DETAIL]). Calm and practical. Under 140 words.
8. The post-hearing summary
Format my notes into a post-hearing summary for [Client]: [PASTE ATTORNEY NOTES]. Structure: what happened, what the court ordered or scheduled, what happens next, what [Client] needs to do. Do not add interpretation or advice beyond the notes. ATTORNEY REVIEW REQUIRED. Under 180 words.
9. The "nothing new" check-in
Write a brief status check-in to [Client] for a matter with no new developments: one line confirming nothing has changed since [LAST UPDATE], one line on what we're waiting for ([NEXT EVENT]), and reassurance that we'll contact them the moment there's news. The silence-filler that prevents anxious calls. Under 70 words.
Settlement, ADR & post-judgment (3 prompts)
10. The settlement offer breakdown
Format this settlement offer summary into plain English for [Client]: [PASTE THE TERMS, GENERICIZED]. Structure by property, support, and parenting terms. Neutral presentation — no recommendation to accept or reject; that's our conversation, not this document. ATTORNEY REVIEW REQUIRED. Under 200 words.
11. The mediation process explainer
Write an explainer for [Client] preparing for their first mediation session: what mediation is and isn't, the mediator's role, roughly how a session flows, and how decisions get made (by the parties, not the mediator). Reassuring, realistic. Under 160 words.
12. The post-judgment compliance check-in
Write a check-in to [Client] [TIMEFRAME] after final judgment: how the [ARRANGEMENT, e.g., parenting schedule] is working, a reminder of what to document if issues arise (dates, specifics, written communication), and when to call us versus when to let small friction pass. Under 130 words.
Save yourself the setup
The 12 prompts above cover the highest-frequency writing in a family law practice — a fraction of a full caseload's workload. If you'd rather skip straight to a tested, privilege-aware library, that's what the 100 ChatGPT Prompts for Family Law Attorneys pack is built for: 104 prompts across custody and parenting-plan communication, financial discovery, property division and settlement communication, court proceedings and case-status updates, high-conflict client communication, mediation and ADR, post-judgment practice, and a privilege-safe AI-use compliance toolkit. (Looking for the consultation-to-signed-client side instead? Our Law Firm Intake & Client Communication pack covers that — the two are built to work together.) Get it at thedailystackstudio.gumroad.com/l/wtxnw.
Frequently Asked Questions
1. Are ChatGPT conversations protected by attorney-client privilege? Treat them as not privileged and potentially discoverable. That's the entire reason for the placeholder-first workflow: nothing identifying ever enters the tool, so there's nothing sensitive to discover.
2. Can AI draft legal documents for my family law cases? This article and pack cover client communication — explainers, updates, check-ins. Pleadings, agreements, and anything filed with a court are attorney work product that demands attorney drafting and verification; AI formatting help there requires extreme care and full review.
3. What does ABA Formal Opinion 512 say? The 2024 opinion addresses lawyers' use of generative AI, emphasizing competence, confidentiality (informed consent before inputting client information into tools that may share or learn from it), supervision, and candor. The placeholder-first workflow is designed around its confidentiality core.
4. Can my paralegal send AI-drafted client updates directly? Logistics messages from reviewed templates, generally yes under supervision. Anything substantive — rulings, settlement terms, strategy-adjacent content — goes through the supervising attorney first (Model Rule 5.3).
5. Is it safe to put financial details into prompts during discovery work?
Keep account numbers, balances, and institution names out. Use [FIGURE]/[ACCOUNT TYPE] placeholders and work with real numbers only inside your case management system.
Conclusion
None of this replaces your judgment, your strategy, or your duty of confidentiality — it replaces the blank cursor before the hardest email of your week. Draft with placeholders, keep strategy out of the chat window, and route every substantive draft through attorney review. Start with the three case-status prompts above — the "nothing new" check-in alone prevents more anxious client calls than anything else on this list — and if they earn a place in your practice, the full 100 ChatGPT Prompts for Family Law Attorneys pack (thedailystackstudio.gumroad.com/l/wtxnw) turns it into a system for the whole office, privilege-safe from the first draft to the last.