The Truth About Using ChatGPT for Family Court as a Litigant in Person

Part 1 of the JSH Law Legal Tech & AI for Litigants in Person Series

By Jessica Susan Hill | JSH Law

The legal world is changing faster than most people realise.

While politicians continue debating court reform, legal aid and access to justice, millions of ordinary people have already started turning to artificial intelligence for help with legal problems. Increasingly, the first “person” a litigant in person speaks to is not a solicitor, barrister, Citizens Advice adviser or legal helpline.

It is ChatGPT.

That reality is now impossible to ignore.

The Master of the Rolls, Sir Geoffrey Vos, recently acknowledged that AI is now being used by “almost every individual litigant in person and small business”, noting that the first port of call used to be a lawyer if one was available and affordable — but is now often ChatGPT or Microsoft Copilot instead. :contentReference[oaicite:0]{index=0}

This article is not anti-AI.

Nor is it blind optimism dressed up as innovation.

This is a practical, honest and experience-based examination of what ChatGPT can genuinely do for litigants in person in family court, where it helps, where it becomes dangerous, and how the justice system itself is now responding to the rapid rise of AI-assisted litigation.

The Reality Facing Litigants in Person

The rise in litigants in person has transformed family court.

Many people now find themselves managing:

• court forms;
• position statements;
• witness statements;
• chronologies;
• Scott schedules;
• evidence bundles;
• emails to the court;
• CAFCASS involvement;
• safeguarding allegations;
• hearing preparation;
• cross-examination;
• procedural deadlines;
• legal terminology;
• and emotionally overwhelming litigation — alone.

Against that backdrop, it is hardly surprising that people are turning to AI tools for assistance.

ChatGPT offers something many litigants in person have never previously had access to:

• instant drafting support;
• plain-English explanations;
• document organisation;
• summarisation;
• issue spotting;
• help structuring arguments;
• and the feeling that someone — or something — is helping them think clearly.

That matters more than many professionals appreciate.

What ChatGPT Actually Does Well

There is a tendency in some legal circles either to dismiss AI entirely or to hype it unrealistically. Both positions are unhelpful.

Used properly, ChatGPT can be extremely useful for litigants in person.

1. Structuring witness statements

One of the biggest problems litigants face is structure.

People often know what happened but struggle to explain it clearly in a chronological, relevant and court-focused way.

ChatGPT is often genuinely helpful at:

• turning chaotic notes into coherent drafts;
• improving readability;
• creating headings;
• removing repetition;
• identifying missing context;
• improving chronology flow.

That alone can significantly improve a litigant’s ability to present their case.

2. Chronologies and timelines

Family court cases frequently involve years of fragmented information spread across:

• WhatsApp messages;
• emails;
• school records;
• police disclosures;
• medical documents;
• social services records;
• CAFCASS reports;
• screenshots;
• and court orders.

ChatGPT is surprisingly good at helping organise information into:

• chronologies;
• event summaries;
• issue lists;
• hearing notes;
• and thematic evidence groupings.

For litigants in person, that can reduce overwhelm enormously.

3. Plain-English explanations

Family court terminology is intimidating.

PD12J, Section 7 reports, threshold criteria, welfare analysis, fact-finding hearings, prohibited steps orders, specific issue applications, directions hearings, final hearings — the language itself becomes a barrier.

ChatGPT is often effective at translating procedural language into plain English.

That improves accessibility.

4. Hearing preparation

Many litigants use ChatGPT to help:

• prepare hearing notes;
• identify key points;
• draft position statements;
• prepare questions;
• organise submissions;
• and rehearse explanations.

Again, used properly, this can genuinely improve confidence and clarity.

What ChatGPT Gets Dangerously Wrong

This is where the conversation becomes critical.

Because while AI can help litigants in person, it can also seriously damage cases if used carelessly.

1. Fake case citations and “hallucinations”

The biggest legal AI risk currently facing the courts is hallucinated authority.

Generative AI systems sometimes invent:

• court cases;
• quotations;
• legal principles;
• articles;
• and procedural rules.

These fabricated answers often look convincing.

That is what makes them dangerous.

The issue has become so serious that the High Court has already issued warnings regarding the misuse of AI-generated legal citations after fake authorities appeared in court proceedings. :contentReference[oaicite:1]{index=1}

The Bar Council’s updated guidance on generative AI specifically warns barristers about hallucinations and entirely false legal information produced by large language models. :contentReference[oaicite:2]{index=2}

Put simply:

2. False confidence

AI systems are designed to sound coherent and confident.

They do not “know” when they are wrong in the way humans do.

That creates a major risk for litigants in person who are already emotionally invested in their case and may mistake confidence for accuracy.

In family court, nuance matters.

Context matters.

Credibility matters.

Human behaviour, safeguarding dynamics and judicial discretion cannot simply be reduced to predictive text.

3. Privacy and confidentiality concerns

Family court cases routinely involve:

• children’s information;
• medical records;
• domestic abuse allegations;
• school information;
• mental health disclosures;
• police material;
• safeguarding concerns.

Litigants in person often upload highly sensitive material into AI systems without understanding:

• how the information is stored;
• how it may be processed;
• whether data may be retained;
• or the confidentiality implications.

The Law Society has repeatedly warned solicitors about confidentiality, privacy and professional responsibility risks connected to generative AI use. :contentReference[oaicite:3]{index=3}

What Guidance Have Lawyers Been Given About AI?

One of the most important developments in the legal world over the past 18 months is that regulators, judges and professional bodies are no longer treating AI as theoretical.

They are actively issuing guidance.

The Law Society

The Law Society has published extensive guidance on generative AI and legal practice, including warnings about:

• hallucinated legal authorities;
• data confidentiality;
• accuracy risks;
• ethical responsibilities;
• supervision obligations;
• professional accountability.

Their guidance can be found here:

Law Society – Generative AI: the essentials

The Bar Council

The Bar Council has issued updated guidance specifically addressing barristers’ use of ChatGPT and other generative AI systems. :contentReference[oaicite:4]{index=4}

The guidance highlights:

• hallucination risks;
• professional integrity obligations;
• confidentiality concerns;
• verification responsibilities;
• and the need for human oversight.

The Bar Council has also acknowledged the growing misuse risks associated with AI-generated legal material. :contentReference[oaicite:5]{index=5}

The Judiciary

The Courts and Tribunals Judiciary issued updated judicial guidance on AI in October 2025. :contentReference[oaicite:6]{index=6}

The judiciary’s guidance recognises both:

• the potential efficiency benefits of AI;
• and the substantial risks associated with accuracy, bias, confidentiality and fabricated material.

Importantly, the judiciary expressly recognises that AI chatbots are now being used by unrepresented litigants. :contentReference[oaicite:7]{index=7}

Are Courts Looking at Restricting AI Use by Litigants in Person?

This is where things become especially interesting.

At present, there is no outright ban preventing litigants in person from using ChatGPT or other AI systems to help prepare court documents.

In reality, such a ban would likely be impossible to police.

However, courts and regulators are increasingly focused on:

• accuracy;
• transparency;
• verification obligations;
• misleading material;
• and abuse of process risks.

The Civil Justice Council established a working group examining the use of AI in preparing court documents. :contentReference[oaicite:8]{index=8}

Current discussions increasingly focus on:

• whether AI-assisted material should be disclosed;
• whether certification requirements may emerge;
• whether professional sanctions should apply for misuse;
• and how courts should manage AI-generated inaccuracies.

There is also growing concern that litigants in person may unknowingly rely on fabricated legal material generated by AI systems.

At the same time, the courts are clearly aware that AI may significantly improve access to justice for people who cannot afford legal representation.

This creates a difficult tension:

The Access to Justice Argument

This is the part of the conversation many commentators miss.

It is easy to criticise litigants in person for using ChatGPT.

But many people are using AI because they genuinely have no realistic alternative.

Legal aid has dramatically narrowed.

Family court representation is expensive.

Court processes are procedurally complex.

Many litigants are traumatised, exhausted and overwhelmed.

Against that backdrop, AI tools can provide:

• structure;
• clarity;
• organisation;
• confidence;
• and practical drafting support.

The legal profession needs to engage honestly with that reality.

Because whether lawyers approve or not, litigants in person are already using AI at scale.

How I Believe ChatGPT Should Be Used in Family Court

Used responsibly, ChatGPT can be extremely powerful as a support tool.

My view is that litigants in person should use AI for:

• document organisation;
• chronologies;
• draft structuring;
• summarisation;
• plain-English explanations;
• hearing preparation;
• question preparation;
• issue identification;
• evidence management.

But not as:

• a replacement for legal judgment;
• a source of unchecked legal authority;
• a substitute for safeguarding analysis;
• a substitute for professional legal advice where needed;
• or a machine that can “win” family court.

The Future of Family Court and AI

The direction of travel is obvious.

AI is not leaving the justice system.

The real questions are now:

• how responsibly it will be used;
• how courts will regulate it;
• how litigants in person will rely upon it;
• how professionals will supervise it;
• and whether it will improve or damage access to justice.

In the next few years we are likely to see:

• AI-assisted bundle preparation;
• AI chronology generation;
• AI hearing summaries;
• AI evidence categorisation;
• AI-assisted legal triage;
• and increasingly sophisticated legal workflow systems.

But the human element will remain critical.

Family court cases involve:

• children;
• risk;
• credibility;
• emotion;
• trauma;
• relationships;
• safeguarding;
• judicial discretion.

No language model truly understands those things.

At least not yet.

Final Thoughts

ChatGPT is neither the saviour of access to justice nor the collapse of the legal profession.

It is a tool.

A very powerful one.

Used properly, it can help litigants in person reduce overwhelm, improve structure and prepare more effectively.

Used badly, it can introduce inaccuracies, fabricated authorities, procedural confusion and false confidence.

The future will belong not to the people who reject AI entirely, but to those who learn how to use it critically, ethically and intelligently.

The legal profession is already adapting.

The courts are already responding.

And litigants in person are already using these tools whether the justice system is ready or not.

Part of the JSH Law Legal Tech & AI for Litigants in Person Series

• Part 1 — ChatGPT for Litigants in Person
• Coming next: Advicenow Review for Litigants in Person
• Coming next: CourtNav and the Rise of Digital Injunction Applications
• Coming next: Claude vs ChatGPT for Family Court Preparation

Regulatory & Editorial Notice: JSH Law Ltd provides litigation support and McKenzie Friend services to litigants in person. JSH Law Ltd is not authorised or regulated by the Solicitors Regulation Authority. This article is provided for general educational and public-interest purposes only and does not constitute legal advice. Court users remain responsible for the accuracy of documents filed with the court and should seek advice from a suitably qualified regulated legal professional where necessary.