Tag Archive for: Child Participation

Family Justice, Recoded?  |  September 2026

Eight developments in family justice — and a practical test for whether reform is reaching the people it is supposed to protect.

JSH Law analysis  •  Family justice, evidence and legal technology  •  England and Wales

A safeguarding concern becomes a paragraph in a report. A relationship becomes a contact arrangement. A frightened parent becomes “difficult to engage”. A child’s account becomes a sentence beginning, “Wishes and feelings were obtained.”

Each description may be accurate. Each may also leave something important out.

My concern is what happens between the original experience and the decision made about it: the context that disappears, the uncertainty that hardens into fact, and the recommendation that reaches an order without anyone clearly explaining how it got there.

That is where I believe the next serious conversation about family justice belongs.

Across domestic-abuse practice, court transparency, artificial intelligence, expert evidence, neurodiversity, adoption, Child Focused Courts and care, I keep returning to the same question:

Earlier editions of Family Justice, Recoded? examined the distance between the lawyer’s desk and the litigant’s kitchen table, and the dangers of reading digital evidence without its context.

This article takes that argument into the everyday work of family justice.

For parents, this article also introduces the JSH Law Six-Question Check: a practical framework for examining important conclusions through Source, Status, Context, Participation, Consequence and Responsibility. For practitioners, policymakers and technology teams, there is a wider challenge: make the reasoning, limitations and responsibility visible.

The legal and procedural focus is England and Wales. The care-leaver development discussed below applies to England. Sources were checked on 27 September 2026.

1. Domestic abuse: preserve the pattern when organising the evidence

A well-organised file can still tell the wrong story if its structure removes the context that gives events meaning.

Take a fictional example.

A parent submits three messages about money. Read separately, they look like disagreements about household expenditure. Read alongside restrictions on access to an account, repeated demands for receipts and consequences for spending without permission, they may require a different inquiry.

The pattern must be evidenced. It cannot simply be asserted. But the method of presentation should allow the pattern to be examined.

Practice Direction 12J, paragraphs 16–19, requires early consideration of whether disputed allegations need determination for risk assessment and welfare decisions.

A separate fact-finding hearing is not automatic. Paragraph 19 expressly accommodates patterns of coercive and controlling behaviour and a holistic examination of alleged abuse.

There is also a concrete 2026 practice development. The President of the Family Division’s revised non-molestation guidance took effect on 12 January 2026, accompanied by Family Justice Council guidance containing a model witness statement.

My practical recommendation is to connect each material allegation to its source, context, impact and relevance to the decision sought.

Preserve the other party’s response and contradictory material too.

2. Transparency: scrutiny depends on preserving the status of the evidence

The open reporting provisions allow accredited journalists and legal bloggers to report children proceedings within the applicable transparency framework, with safeguards for privacy and anonymity.

That is an important opportunity for public understanding.

It also demands precision.

An allegation, an admission, a professional opinion and a judicial finding have different evidential status.

A report that blurs them can leave readers with a materially inaccurate understanding of the case.

The same applies inside the court file.

If a disputed account is repeated through several documents, ask whether those documents provide independent support or whether they are repeating the same original source.

My proposed discipline for public commentary is simple:

• identify what the court was deciding;
• identify what the court actually found;
• identify what remains disputed; and
• identify what the available record cannot establish.

3. AI: every important sentence needs a route back to its source

AI can produce an orderly account from disorderly documents.

That is attractive when someone has years of messages, several reports and a hearing approaching.

The risk is that the orderliness creates confidence the underlying work has not earned.

An approximate date becomes exact.

“She reported” disappears.

An expert’s qualification is omitted.

A disputed allegation becomes a factual chronology entry.

The document reads better while becoming less reliable.

The Solicitors Regulation Authority’s warning notice on misuse of AI addresses inaccurate material and confidentiality risks.

It makes clear that regulated professionals remain accountable and that paid tools, as well as free tools, may lack suitable safeguards.

Separately, the Civil Justice Council’s June 2026 update on the use of AI in preparing court documents identifies witness-statement authenticity as an area requiring further consideration.

That work should not be described as a new binding family-court rule.

A practical standard for AI-assisted case preparation

• Every substantive factual entry should be traceable to the original material.
• Uncertainty and disagreement should remain visible.
• A witness should recognise and understand their own account.
• Legal authorities should be opened and checked.
• Sensitive information should enter a system only after its suitability has been assessed.
• An identifiable person should review the final work.

These are proposed working controls, not a claim that any particular product guarantees reliability.

4. Toxicology: keep the result, interpretation and welfare decision distinct

An expert report may carry enormous weight.

Its authority should make its reasoning easier to inspect, not harder.

The Family Justice Council’s hair-strand-testing working group identifies concerns about limitations, interpretation and the weight placed on reports.

Its published remit includes factors affecting interpretation and the quality of instructions.

The page describes work towards guidance; it should not itself be represented as completed guidance.

Family Procedure Rules Part 25 sets out the expert’s overriding duty to assist the court and includes provisions concerning questions to experts.

Those steps should not merge into one another without explanation.

Useful questions include:

• what period does the result address?
• what limitations does the report record?
• what relevant information was supplied to the expert?
• what assumptions were made?
• does the report answer the question the court actually needs resolved?

A parent disputing a result needs focused clarification or appropriate expert advice.

A generic internet argument about testing will rarely identify the issue in their particular report.

5. Neurodiversity: identify the barrier before judging the response

Suppose a person gives a confused answer after being directed to an unfamiliar page and asked several questions at once.

Before drawing conclusions from that answer, I would want to know whether they could:

• find the relevant passage;
• read it;
• process the questions being asked;
• retain the information long enough to answer; and
• understand what was actually being asked of them.

The Family Justice Council’s neurodiversity guidance encourages attention to individual needs and barriers.

Family Procedure Rules Part 3A requires consideration of participation and evidence where vulnerability is relevant.

My practical recommendation is to describe the difficulty and the adjustment together.

• “I need time to read the passage before answering.”
• “Please ask one question at a time.”
• “I need written confirmation of the directions.”
• “I need planned breaks to maintain concentration.”

The appropriate arrangements depend on the individual and the court’s decisions.

They should be discussed early and checked in practice.

This also changes how I would review descriptions such as “uncooperative”.

What behaviour supports the label?

What alternative explanation was considered?

What support or adjustment was available?

6. Post-adoption contact: a plan needs people and resources behind it

“Letterbox contact” can sound like a complete answer.

It leaves practical questions.

Who helps write?

Who follows up?

What happens when correspondence stops?

How are the child’s changing needs considered?

The Public Law Working Group’s November 2024 adoption recommendations call for significant improvement, including greater consideration of face-to-face relationships where appropriate and stronger support.

Those recommendations do not create an automatic entitlement to direct contact.

Individual needs and safety remain central.

My concern is implementation.

A proposal may be carefully worded but depend on support nobody has agreed to provide.

For each important relationship, ask:

• What does this relationship mean to this child?
• What benefits and risks have been assessed?
• What arrangement is proposed?
• Who will make it happen?
• What support is needed?
• How will difficulties be addressed?
• How will the arrangement respond to the child’s changing needs?

The answers may differ for a birth parent, sibling or grandparent.

Those relationships deserve individual consideration.

Digital communication adds further questions about privacy, boundaries and support.

A new platform cannot answer them simply by making messages easier to send.

7. Child Focused Courts: measure what happens after the child is heard

The Government’s March 2026 announcement sets out national expansion of Child Focused Courts, previously known as Pathfinder.

It is a rollout programme, not proof that every court or every existing case already follows the model.

The January 2026 evaluation of family experiences provides valuable evidence from Dorset and North Wales.

It involved 39 parents and carers and nine children, with self-selecting participants.

Its findings should not be turned into a nationally representative success rate.

My test for meaningful participation has five parts:

1. a suitable opportunity for the child to communicate;
2. an accurate record of what was communicated;
3. consideration of the context;
4. an explanation of how the child’s account informed the recommendation; and
5. an understandable explanation of the outcome.

That goes further than recording that wishes and feelings were obtained.

It also avoids making the child responsible for choosing the outcome.

For adults preparing a response to a report, identify the type of concern.

Is it:

• a factual error?
• a missing source?
• an unexplained inference?
• a participation problem?
• or simply disagreement with the recommendation?

Different problems require different responses.

8. Growing up in care: name who will still be there

Children experience care through relationships and daily life.

A completed review cannot, by itself, tell us whether a child has someone dependable to call.

The Children’s Commissioner’s work on siblings in care illustrates the importance of relationships and the practical difficulties of maintaining them.

That report dates from 2023; its figures should not be presented as current 2026 prevalence data.

There is also an immediate development in England.

Revised care-leaver local-offer guidance takes effect on 30 September 2026.

At this article’s research date, 27 September 2026, that date is still forthcoming.

The guidance covers accessible information about accommodation, support and relationships, while distinguishing legal entitlements from additional local provision.

That question belongs beside every outstanding action concerning housing, education, health and important relationships.

Where an arrangement fails, the record should show the next responsible person rather than leave the young person to restart the search for help.

The JSH Law Six-Question Check

One practical way to examine the reasoning in family-court material is to take one important conclusion, recommendation or factual proposition and ask six questions.

The framework does not decide whether an allegation is true, whether a professional recommendation is correct or what order the court should make. It is a preparation tool designed to make the evidential and reasoning chain easier to examine.

What I want family justice to make visible

Across these eight areas, my proposal is the same:

preserve a clear route from source material to analysis, from analysis to decision, and from decision to action.

That gives a parent something specific to understand or challenge.

It gives a practitioner a way to identify gaps.

It gives a technology team a meaningful design requirement.

And it gives a child a better chance of seeing that their experience affected what happened next.

My work brings together evidence organisation, family-court preparation and digital service design.

I am interested in the points where people lose the ability to understand or influence a process — and in practical ways to restore it.

For anyone commissioning reform or legal technology, I would ask for a demonstration using a difficult example:

• a disputed date;
• an incomplete record;
• a vulnerable witness;
• an allegation with competing accounts; or
• an expert opinion containing important qualifications.

Show how the system preserves that complexity.

Show how a user corrects an error.

Show what other conclusions are affected when that correction is made.

And show who remains responsible when the software stops and the real-world decision begins.