Trauma-Informed Lawyering Is Not Enough: Legal Technology Needs Guardrails Too

I recently read a LinkedIn post by Christine Dunn, Washington DC Managing Partner at Sanford Heisler Sharp McKnight and Co-Chair of its Sexual Violence, Title IX and Victims’ Rights Practice Group.

Christine described seeing survivors treated by lawyers and judges in ways that intensified their trauma. Her argument was direct: trauma-informed training should be mandatory for law students because misunderstanding trauma can distort how hesitation, fragmented recall and behaviour are interpreted.

She was discussing a newly published research article: Trauma-Informed Lawyering: Training as a Foundation for Trauma-Responsive Courtrooms and Legal Practices, by Mona Mittal, Anna Paden McCormick, Bridgette Stumpf, Lindsey Silverberg and Gabriel Dogbanya.

I agree with Christine’s central point. But I would take it one stage further.

Trauma-informed practice should not live only inside a lawyer’s training. It should be built into the legal systems, court processes and technology that shape what the lawyer does next.

Legal technology is already helping professionals conduct intake, draft correspondence, organise evidence, prepare witness statements, assess cases, manage deadlines and communicate with clients. If those systems reward speed, certainty and standardisation without understanding trauma, they can industrialise the very behaviours trauma-informed training is supposed to prevent.

The better future is not an AI system that labels a person “traumatised”. It is technology that gives the human professional timely, evidence-safe guardrails: explain why sensitive information is needed; offer a real choice where one exists; check safe methods of contact; avoid unnecessary repetition; identify a possible participation issue; preserve uncertainty in a witness account; and require a trained human to make the judgment.

In this article

• What the new research does—and does not—show.
• What trauma-informed lawyering actually means.
• Why legal processes can unintentionally re-traumatise people.
• What the law in England and Wales already requires.
• Why professional duties have not yet produced consistently trauma-responsive practice.
• Seven guardrails that should be built into legal technology.
• The red lines: what legal AI must never be allowed to do.
• Practical steps for litigants in person whose participation is affected.

What the new trauma-informed lawyering study found

The research article was published online in July 2026 in the Journal of Aggression, Maltreatment & Trauma. According to its publicly available abstract, the authors examined an innovative 90-minute trauma-informed training session for lawyers.

Four focus-group discussions took place between March and September 2024 with a total of 15 legal professionals who had attended the training. The thematic analysis identified several reported effects:

• participants described the training as transformative;
• many identified a significant gap in conventional legal education;
• participants reported gaining practical communication skills for working with clients who had trauma histories;
• they reported becoming better able to recognise reactions and behaviours that may be connected with traumatic stress;
• they developed greater awareness of vicarious trauma in their own professional lives; and
• they wanted further training, particularly on vicarious trauma.

The authors conclude that trauma-informed lawyering training may strengthen lawyer-client relationships, reduce inadvertent re-traumatisation and support lawyer wellbeing.

That is promising. It is not, however, proof that a single 90-minute session transforms legal outcomes. The study used qualitative data from a small, self-selected group of 15 participants. The authors expressly say that the findings are preliminary and are not intended to be generalisable. Participants reported changes in understanding and practice; the abstract does not establish long-term behavioural change, improved case outcomes or reduced re-traumatisation measured from clients’ own perspectives.

The right response is therefore neither to dismiss the study because it is small nor to oversell it because its conclusion is attractive. It should be treated as an important early signal and a case for better, larger and longer-term research—including research led by the people who use legal services.

What trauma-informed practice actually means

The UK Government’s working definition of trauma-informed practice was developed for health and care settings, but its framework is highly relevant to justice.

It identifies six principles:

1. Safety
2. Trust
3. Choice
4. Collaboration
5. Empowerment
6. Cultural consideration

The purpose is not to provide trauma treatment. Lawyers are not therapists and legal software is not a clinical service. The purpose is to recognise how trauma can create barriers to engaging with a service and to reduce avoidable re-traumatisation.

In legal work, that can mean telling a client why a painful question is necessary before asking it. It can mean obtaining the essential account without demanding that the person repeat it to five different members of staff. It can mean explaining the next procedural step instead of allowing silence and uncertainty to do further damage. It can mean asking what helps the person understand and communicate, rather than assuming that one standard appointment, form or hearing format works for everyone.

Trauma-informed does not mean trauma-assuming

This distinction matters.

Trauma can affect attention, recall, emotional regulation, trust and communication. A person may remember sensory fragments before dates. They may become silent, detached, highly distressed or apparently calm. They may disclose incrementally. Equally, none of those behaviours proves that trauma occurred, and there is no single presentation shared by all survivors.

An inconsistency is something to examine. It is not, by itself, a diagnosis or a verdict. Trauma may be relevant; so may ordinary memory fallibility, the passage of time, fear, shame, language, disability, neurodivergence, misunderstanding, suggestion or fabrication. The legal task remains to test the evidence fairly.

Using trauma to explain every inconsistency would be as unsafe as using every inconsistency to disprove trauma.

A trauma-informed process does not predetermine the answer. It improves the conditions in which instructions can be obtained, evidence can be given and disputed facts can be tested. Proper challenge may still be robust. It does not need to be needlessly humiliating, confusing or destabilising.

How legal processes can intensify trauma

The legal system routinely asks people to perform difficult cognitive tasks at the worst point in their lives.

A person may be expected to:

• produce a precise chronology while living in fear or crisis;
• repeat distressing events to police, solicitors, support services, Cafcass, experts and the court;
• understand procedural language while sleep-deprived or hypervigilant;
• read hostile allegations and respond within a short deadline;
• sit near, see or be questioned about an alleged perpetrator;
• navigate digital portals designed around administrative efficiency rather than emotional safety;
• make important decisions without knowing what will happen next; and
• remain calm enough to be seen as credible while describing events that did not feel calm at all.

For litigants in person, there is an added problem. The person must be witness, case manager, document reviewer, procedural researcher and advocate at the same time. The system can then interpret the resulting disorganisation as a personal failing instead of asking whether the process itself has reduced the person’s ability to participate.

That does not make every distressed litigant right about the merits of the case. It does mean that avoidable procedural harm can reduce the quality of information available to the court.

England and Wales already recognises vulnerable participation

Trauma-informed lawyering is not an express, universal legal code in England and Wales. But important parts of the existing framework already point in the same direction.

Family Procedure Rules Part 3A and Practice Direction 3AA

Part 3A of the Family Procedure Rules requires the court to consider whether vulnerability is likely to diminish a party’s participation or the quality of a party’s or witness’s evidence and, if so, whether participation directions are necessary.

Where it is stated that a party or witness is, or is at risk of being, a victim of domestic abuse by a relevant person, rule 3A.2A requires the court—unless the person asks otherwise—to apply the prescribed assumption for the purpose of considering participation directions.

Practice Direction 3AA requires vulnerability to be identified at the earliest possible stage. Measures may include preventing one person from seeing another, live-link participation, communication assistance, an intermediary and other tailored directions. A ground-rules hearing may be used to decide how questioning and participation should be managed.

This is a useful model for trauma-informed legal technology: identify a possible barrier, ask what practical support may help, preserve the person’s own views and leave the factual and legal decision to the proper human decision-maker.

Judicial guidance and professional duties

The Judicial College’s Equal Treatment Bench Book helps judges identify circumstances affecting communication and participation and consider appropriate adjustments.

The SRA’s current framework also contains relevant duties. Paragraphs 3.2 to 3.4 of the Code of Conduct for Solicitors require competent service, maintenance of competence and consideration of a client’s attributes, needs and circumstances. The Statement of Solicitor Competence requires solicitors to respond sensitively to individual characteristics and take reasonable steps to meet the service needs of clients in vulnerable circumstances.

The SRA’s own guidance says services may need anticipatory adjustments, flexible appointment arrangements, adapted communication, specialist support and attention to the person’s communication preferences.

These are meaningful obligations. They are not the same as a universal, express requirement for every law student and every practising lawyer to complete trauma-informed lawyering training. The current SQE framework addresses vulnerable clients in some contexts, but it does not prescribe a comprehensive trauma-informed practice module across legal education.

The implementation gap

Rules and professional principles matter, but they do not automatically change the moment-to-moment conduct of legal work.

A lawyer may complete annual training and then return to a case-management system that:

• forces a client to repeat the same account at every stage;
• sends sensitive notifications without checking whether the device or inbox is safe;
• rewards staff for closing an intake quickly rather than understanding a communication need;
• turns uncertainty into a polished but falsely confident narrative;
• flags a distressed client as “difficult” or “non-compliant”;
• records no safe-contact preference;
• fails to prompt consideration of Part 3A or PD3AA; or
• uses AI-generated correspondence without checking its tone, accuracy or effect.

That is why legal technology should not be treated as neutral infrastructure. Its forms, defaults, labels, reminders, performance measures and automated suggestions influence professional behaviour.

Seven guardrails for trauma-informed legal technology

The six recognised trauma-informed principles can be translated into concrete product requirements. Legal technology also needs a seventh principle: accountable human judgment.

The lawyer-facing guardrail: a pause before pressing send

Trauma-informed legal tech should not merely change the client-facing screen. It should help the lawyer reflect at the point of action.

Before a sensitive request or letter is sent, the system could prompt:

• Have you explained why this information is needed?
• Is the language clear and firm, or accusatory and shaming?
• Does the client need to provide all of this now, or can the request be staged?
• Are you asking the client to repeat information already held?
• Is the deadline imposed by the court or created by your workflow?
• Have you told the client what will happen after they respond?
• Have safe contact preferences been checked?
• Does the draft distinguish allegation, evidence, professional opinion and legal conclusion?
• Has a human verified the facts, law, tone and intended recipient?

This is not a software veto over professional judgment. It is the digital equivalent of a good supervisor asking the right question before an avoidable mistake is made.

Evidence tools must preserve uncertainty rather than erase it

Generative AI is very good at turning disordered information into fluent prose. In trauma-related cases, that strength can become a serious evidential weakness.

A client may say:

“I cannot remember whether that message came before or after I went to my sister’s house. I know it was the same week.”

An unsafe system may quietly choose a date, place the events in an attractive chronology and remove the uncertainty. The resulting statement sounds stronger. It is also less faithful to the witness’s evidence.

A trauma-informed and evidence-safe system should:

• retain the person’s original words or source recording where lawful and appropriate;
• mark uncertainty explicitly;
• separate direct recollection from inference and information received from others;
• flag an inconsistency neutrally for human clarification;
• show every material amendment to the person approving the statement;
• never manufacture precision, emotion or certainty; and
• preserve version history and an audit trail.

The goal is not to make an account look perfect. It is to make the record accurate, intelligible and properly attributable.

The red lines: what legal AI must not do

Some uses are too risky to be dressed up as trauma-informed innovation.

Legal systems should not:

• diagnose PTSD or any other condition from a client’s words, voice, face or behaviour;
• produce a “trauma score”, “victim score” or credibility rating;
• treat emotional presentation as evidence that an allegation is true or false;
• recommend child arrangements, bail, sentence, liability or case outcome from a trauma label;
• classify a person as difficult, manipulative or non-compliant without contextual and human review;
• rewrite a witness’s evidence without transparent confirmation;
• infer sensitive health information merely because it may be commercially useful;
• monitor lawyers’ communications to create hidden mental-health or resilience scores; or
• use a compassionate interface to disguise weak legal accuracy, poor security or the absence of human accountability.

A calm colour palette is not trauma-informed design. The test is what the system does with power, information, choice and error.

Confidentiality, privacy and automated decisions

Trauma-related legal material may include health information, sexual-life data, allegations of crime, children’s information, addresses, immigration status and highly sensitive family evidence. A system that infers a mental-health condition may itself be creating special-category personal data.

In August 2026, the SRA issued a specific warning notice on misuse of AI. It warned that entering client-sensitive information into open or unsuitable AI systems can breach confidentiality and place legal professional privilege at risk. Paid access alone does not make a system safe; the provider’s terms, settings, retention, architecture and use of data must be understood.

Legal organisations therefore need, at minimum:

• a lawful basis and, where required, an additional condition for sensitive processing;
• data minimisation and strict purpose limitation;
• clear retention and deletion rules;
• appropriate security and access controls;
• privacy information people can actually understand;
• a data-protection impact assessment for high-risk processing;
• controls preventing confidential material from being used to train an external model without proper authority; and
• meaningful routes to correction, challenge and human intervention.

Following the Data (Use and Access) Act 2025, the UK GDPR’s current automated-decision framework is contained in Articles 22A to 22D. Where a solely automated decision has a legal or similarly significant effect, the statutory safeguards include information, the ability to make representations and challenge the decision, and access to human intervention. Those are minimum legal safeguards in relevant cases, not a complete ethical standard for trauma-informed justice.

For related JSH Law analysis, see Before You Press Submit: Why AI Prompt Governance Matters in Legal Work and Family Justice and The JSH Law Legal Tech Test.

Vicarious trauma is an organisational issue—not a resilience defect

The new study’s participants wanted more support on vicarious trauma. That matters because lawyers, caseworkers, judges, court staff and support providers repeatedly read and hear accounts of violence, abuse, bereavement and harm.

LawCare describes vicarious trauma as the range of negative effects that can arise from day-to-day immersion in traumatic material. Effective responses include reflective space, peer and supervisory support, reasonable workload design and access to professional help.

Technology can help by identifying cumulative caseload exposure at a team level, protecting debrief time, prompting supervision after particularly difficult work and making support easy to access. It must not become another surveillance system that secretly analyses private communications, diagnoses staff or penalises people for showing ordinary human responses to traumatic work.

If a workplace repeatedly exposes people to traumatic material, the answer cannot be to tell each individual to become tougher.

What this means for litigants in person in family court

If trauma, domestic abuse, disability or another vulnerability is affecting your ability to participate or give evidence, raise the practical problem as early and specifically as possible. Do not assume that the court will identify it from the papers.

A useful request focuses on function and the measure sought:

Depending on the facts and what is available, practical measures may include:

• screens or arrangements preventing direct sight;
• live-link participation or evidence;
• separate waiting or arrival arrangements;
• scheduled breaks;
• questions asked in shorter, clearer language;
• a ground-rules hearing before evidence is given;
• communication assistance or an intermediary where appropriate;
• attendance by an IDVA or ISVA in accordance with the applicable rules; and
• other directions tailored to the person and the hearing.

The court decides what is necessary and available. A diagnosis is not a substitute for explaining the practical barrier, and a request for participation support is not a finding on the allegations.

If proceedings are already underway, check the latest order and the relevant court procedure before filing anything. Urgent safeguarding concerns and imminent hearings require prompt, case-specific advice where available.

What law firms, courts and developers should do next

A credible trauma-informed programme should reach beyond a single training session.

1. Make foundation training routine. Trauma, communication, cultural context, vicarious trauma and the limits of the lawyer’s role should be built into legal education, induction, advocacy training and continuing competence.
2. Turn principles into workflow requirements. Safe contact, explanation, client confirmation, participation prompts, evidence provenance and human review should be product features, not optional good intentions.
3. Co-design with affected users. Survivors, litigants in person, disabled court users, children’s specialists, defence practitioners, advocates and support services should all be involved. Trauma-informed must not become a one-sided credibility framework.
4. Test real harms. User testing should examine distress, abandonment, disclosure safety, unequal error, misunderstanding, over-reliance and the effect of the product on evidence—not merely completion time and customer satisfaction.
5. Procure for safety and participation. Courts and firms should require evidence of privacy, accessibility, contestability, auditability, safe escalation and trauma-informed design before buying a system.
6. Support the workforce. Supervision, protected reflection, workload management and access to independent support should be organisational responsibilities.
7. Keep humans answerable. A named professional or public authority must remain responsible for consequential legal, evidential and safeguarding decisions.

My view: legal technology should help the profession become more human

The strongest argument for trauma-informed legal technology is not that a machine can understand trauma better than a person.

It cannot.

The argument is that well-designed systems can help human beings remember what pressure, time and professional habit sometimes cause them to forget. Explain. Ask only what is necessary. Preserve the person’s words. Do not convert uncertainty into false precision. Check whether contact is safe. Consider participation early. Make the next step visible. Pause before sending. Escalate to a human.

Trauma-informed practice is not softness. It is disciplined communication, safer evidence-handling and fairer participation.

Nor is it a substitute for legal analysis. The court must still determine facts, apply the burden and standard of proof, protect children and other vulnerable people, and ensure fairness to every party.

The purpose of trauma-informed justice is not to decide the case in advance. It is to stop the process itself from making a person less able to take part in the decision.

Frequently asked questions

What is trauma-informed lawyering?

Trauma-informed lawyering is an approach that recognises how trauma may affect communication, trust, recall, behaviour and participation. The lawyer adapts the service to improve safety, clarity, choice and effective involvement while remaining within the legal and professional role.

Does trauma-informed practice mean believing every allegation?

No. It concerns the fairness and accessibility of the process. It does not prove that trauma or abuse occurred, remove the need for evidence, prevent proper challenge or change the applicable burden and standard of proof.

Can trauma affect memory?

Trauma may affect attention, encoding, retrieval and the order or manner in which an account is given. There is no universal trauma presentation, however, and an inconsistency can have many possible explanations. It must be explored fairly rather than automatically accepted or condemned.

What are participation directions in family court?

Participation directions are case-management measures intended to help a vulnerable party or witness participate or give evidence effectively. Under FPR Part 3A, they may include screens, live link, communication assistance, an intermediary or other tailored arrangements.

Is trauma-informed training mandatory for every lawyer in England and Wales?

There is no single universal requirement that every law student and practising lawyer complete a comprehensive trauma-informed lawyering course. Solicitors nevertheless have duties of competence, communication and attention to client needs, and specific practice areas contain additional vulnerability requirements and guidance.

What is trauma-informed legal technology?

It is legal technology designed to reduce avoidable re-traumatisation and improve effective participation through safe contact, clear explanation, meaningful choice, accessible communication, evidence provenance, privacy, routes to correction and accountable human review.

Should AI identify whether a client is traumatised or truthful?

No. AI should not diagnose trauma, infer credibility from behaviour or generate a victim score. It may support a human-led checklist about communication and participation needs, but consequential clinical, legal, evidential and safeguarding judgments require properly qualified and accountable people.

How can legal technology help lawyers become more trauma-informed?

It can prompt lawyers to explain why sensitive information is needed, check safe contact preferences, avoid unnecessary repetition, use clear language, preserve uncertainty in evidence, consider participation measures, verify AI outputs and pause before sending correspondence likely to affect a vulnerable client.

Need help preparing to participate effectively in family proceedings?

Sources and further reading

• Christine Dunn, LinkedIn post on trauma-informed lawyering and legal education.
• Mona Mittal, Anna Paden McCormick, Bridgette Stumpf, Lindsey Silverberg and Gabriel Dogbanya, Trauma-Informed Lawyering: Training as a Foundation for Trauma-Responsive Courtrooms and Legal Practices, Journal of Aggression, Maltreatment & Trauma (published online July 2026).
• Office for Health Improvement and Disparities, Working definition of trauma-informed practice.
• Ministry of Justice, Family Procedure Rules Part 3A: Vulnerable persons.
• Ministry of Justice, Practice Direction 3AA: Participation in proceedings and giving evidence.
• Judicial College, Equal Treatment Bench Book, July 2024 edition updated February 2026.
• Solicitors Regulation Authority, Meeting the needs of vulnerable people.
• Solicitors Regulation Authority, Statement of solicitor competence.
• Solicitors Regulation Authority, Misuse of AI: warning notice, 17 August 2026.
• Information Commissioner’s Office, Guidance on AI and data protection.
• Department for Science, Innovation and Technology, Data (Use and Access) Act 2025 factsheet: UK GDPR and DPA.
• LawCare, Vicarious trauma in legal work.
• The Advocate’s Gateway, Toolkits on vulnerable witnesses, parties and communication.

About the author

Jessica Susan Hill is the founder of JSH Law. She supports litigants in person with family-court procedure, evidence organisation, safeguarding analysis, court documents and hearing preparation. Her work and study sit at the intersection of family justice, legal technology, artificial intelligence, access to justice and trauma-aware practice.

Regulatory and editorial notice

JSH Law Ltd provides litigation support and McKenzie Friend services to litigants in person. JSH Law Ltd is not a firm of solicitors and is not regulated by the Solicitors Regulation Authority. We do not conduct litigation, do not go on the court record and do not exercise rights of audience unless a court grants permission in a particular hearing.

This article provides general legal information, research commentary and public legal education for England and Wales. It is not legal advice, clinical advice or a substitute for advice about an individual case. Every case turns on its facts, evidence, procedural history, current orders and the directions made by the court.

References to Christine Dunn and the research authors are for attribution and commentary. The views and proposed legal-technology framework in this article are JSH Law’s independent analysis and should not be attributed to those individuals unless expressly stated in their own work.