Human Rights
6
 min read

Early Prisoner Release and Domestic Abuse: The Risk the System Cannot See

After being interviewed by Channel 5 News for its coverage of the Government’s new prisoner-release arrangements, Aneeta Prem MBE examines a wider safeguarding question: whether domestic-abuse risk can be partly invisible to the court and prison record, and what should happen before accelerated release.

Editorial graphic showing an open prison doorway for Aneeta Prem MBE’s analysis on early prisoner release, domestic abuse and victim safety.
Aneeta Prem
September 1, 2026

By Aneeta Prem MBE | 1 September 2026

When Channel 5 News interviewed me today about the early release of prisoners, I kept coming back to one question:

Is the victim safe?

Prison overcrowding is real. Government has to deal with it. Rehabilitation matters. Most prisoners will eventually return to the community.

But none of that changes a basic principle.

We cannot solve a prison-capacity problem by moving the risk from a prison cell to a victim’s front door.

For domestic abuse, that risk deserves particular attention because what happened inside the relationship may be far greater than what appears on the criminal record.

Can domestic-abuse offenders be released earlier?

Yes, some can.

From 1 October 2026, the Government’s new Progression Model changes release arrangements for some people serving Standard Determinate Sentences in England and Wales.

It does not apply to everyone. Life sentences, Extended Determinate Sentences and other more serious sentence types are outside the model. The Government has also excluded rape, specified serious child sexual and grooming offences, unlawful killing, indecent assault and a number of other offences.

Those exclusions are important.

But domestic abuse is not itself a blanket exclusion.

That matters because domestic abuse is not one criminal offence. A perpetrator may be convicted of assault, controlling or coercive behaviour, criminal damage or another offence committed within an abusive relationship.

So simply reading the name of the offence may not tell us enough about the risk to the person waiting outside prison.

The criminal record may show the conviction. The victim knows the history.

This is the heart of my concern.

A conviction tells us what was proved in court.

It does not necessarily tell us everything the victim survived.

Domestic abuse can involve control, threats, intimidation, stalking, financial abuse and violence. Some incidents may never have been reported. Others may have been reported but never charged.

Victims may stay silent because they are frightened of what will happen if they speak. They may depend on the perpetrator financially. They may worry about their children, their home or their immigration status. They may feel ashamed or believe nobody will help. Government statutory guidance recognises many of these barriers.

SafeLives reports that high-risk victims experience domestic abuse for an average of 2.3 years before getting effective help.

That should make us cautious about assuming that the court file contains the whole story.

Silence is not proof of safety. A short criminal record is not proof of a short history of abuse.

Why does that matter when someone leaves prison?

Because behaviour in prison and behaviour towards a particular victim are not necessarily the same thing.

The Progression Model allows prison behaviour to affect how long an offender remains in custody. All offenders covered by the model will serve minimum periods in prison, and poor behaviour can result in longer custody.

There is sense in rewarding genuine rehabilitation.

But in a domestic-abuse case, I would ask another question:

What does behaving well in prison tell us about how somebody will behave towards the person they previously controlled?

A prisoner may comply with rules, staff and routines while still feeling anger, resentment, fixation or a need to regain control over a former partner.

That does not mean every domestic-abuse offender will reoffend. It means prison behaviour alone cannot answer a victim-specific question of risk.

We know the consequences can be fatal

Most domestic abuse does not end in murder. We should never suggest that it does.

But some does.

In the year ending March 2025, there were 111 domestic homicides in England and Wales. Seventy-five victims were women and 36 were men. Sixty-seven people were killed by a partner or former partner.

The same year, the Crime Survey for England and Wales estimated that 3.8 million people aged 16 and over experienced domestic abuse.

These figures are not a reason to frighten people. They are a reason to take risk seriously.

When getting a safeguarding decision wrong can have devastating consequences, we need to understand as much as we reasonably can before accelerating release.

Parliament has already given us part of the answer

The Sentencing Act 2026 contains an important change.

Where a court is sentencing an offender and considers that the offence involved domestic abuse, the court must state that finding in open court. The finding becomes a formal court result.

That gives the justice system a clearer marker saying:

This offence involved domestic abuse.

So we should use it.

My proposal is straightforward: a formal finding of domestic abuse should trigger a specific victim-safety review before accelerated release.

Not a new sentence. Not a veto for the victim. Not an assumption that every offender remains dangerous.

A safeguarding test.

Before accelerated release, decision-makers should know whether there has been stalking, threats, previous violence, breaches, coercive control or attempts to find or contact the victim.

They should know whether children are involved, whether the victim believes the perpetrator still poses a danger and why, whether the offender knows where the victim lives or works, and whether the proposed licence conditions can realistically manage that risk.

If important information is missing, the answer should not be to guess.

Victims must be more than recipients of a release date

The Government has put safeguards into the new arrangements.

People released through the Progression Model will be on licence and subject to supervision. Conditions can include restrictions on contact and movement, and electronic monitoring. Eligible victims in the Victim Contact Scheme can make representations about protective licence conditions, including non-contact conditions and exclusion zones.

I welcome that.

But a victim should not simply be told what is happening after the important decisions have effectively been made.

The victim may hold information that changes the assessment of risk.

That information should be heard early enough to matter.

A tag may tell us where somebody is. An exclusion zone may tell them where they should not go. Neither can replace understanding the perpetrator, the history and the victim’s fear.

This is not an argument against rehabilitation

I want perpetrators to change.

I want prison to do more than contain people until a release date.

And I do not believe that everyone convicted of an offence connected with domestic abuse should automatically be treated as presenting the same level of danger. That would be poor justice.

But good justice also means recognising that risk is not always visible from a sentence length, an offence code or a prison disciplinary record.

Domestic-abuse risk can be intensely personal.

There may be one person outside the prison walls who knows exactly what that offender is capable of.

We should listen.

The test

The Government has a prison-capacity crisis to solve. It also has a duty to protect victims.

Those responsibilities cannot be separated by simply moving somebody from custody into the community.

So where domestic abuse has been formally identified, I believe accelerated release should carry an additional safeguard:

a specific, recorded assessment of the risk to the victim before the earlier release takes place.

That is proportionate. It recognises rehabilitation. It recognises the rights of offenders. And it recognises something the justice system must never forget:

the person leaving prison is not the only person whose life changes when the prison door opens.

The final question should always be:

Is the victim safe?

Sources and further reading

Related on Aneeta.com

Media and professional enquiries

Contact Aneeta Prem

For interviews, expert commentary, speaking and public-interest enquiries.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.