Closing the intimate image abuse gap

Five years ago, parliament passed legislation that many hoped would mark a turning point. The Domestic Abuse Act 2021 made threatening to share intimate images a criminal offence, building on the law introduced six years earlier that had criminalised the act of sharing itself.

Campaigners celebrated. Survivors who had spent years being told by police that nothing could be done finally had a legal framework behind them.

Five years on, new data obtained by the domestic abuse charity Refuge tells a rather different story.

Between July 2021 and February 2026, police forces in England and Wales recorded 21,905 intimate image abuse offences. Of those, just 4.8 percent, 1,047 cases, resulted in a charge or summons. In 56 percent of cases where a suspect had already been identified, nobody was charged at all.

The data, obtained through Freedom of Information requests to all 43 forces in England and Wales, of which 27 responded, with 25 providing a breakdown of outcomes, shows that the problem is not static. It is getting worse.

Recorded offences rose by 26.9 percent between the year ending June 2022 and the year ending June 2025. The number of charges issued over the same period remained essentially flat with 236 perpetrators charged in the year ending June 2022, compared to 233 in the year ending June 2025. In proportional terms, the charging rate fell from 5.8 percent in 2021-22 to 4.5 percent in 2024-25.

The threat to share offence, the specific provision won through Refuge’s campaign, presents a picture that is both encouraging and deeply troubling in equal measure.

Only 11 of the 27 responding forces were able to provide a breakdown of offences specifically recorded as threats to share, which is itself a data quality concern. Among those 11 forces, recorded threats to share offences rose by 344 percent between 2021-22 and 2024-25. Despite this surge, Refuge’s specialist technology-facilitated abuse team reports that in the five years since the legislation came into force, it has not supported a single survivor whose perpetrator has been convicted of threatening to share an intimate image.

Not one conviction. In five years.

Why cases are failing

The Refuge data identifies two principal drivers of case attrition, and both deserve serious attention from policing.

The first is victim withdrawal. In cases where a suspect had been identified, 55.8 per cent, 6,848 cases, did not progress because the victim withdrew or was unable to continue supporting the investigation. This is a significant figure, but it requires careful interpretation. Victim withdrawal in cases of intimate image abuse is rarely straightforward disengagement.

The offence is almost invariably embedded in a broader pattern of coercive and controlling behaviour. Perpetrators frequently use the threat of further disclosure as a mechanism of continued control. A survivor who withdraws from a police investigation may be doing so under pressure, in fear, or because the process of engaging with the criminal justice system has become another means of coercive control.

The second driver is evidential difficulty. Around 25 percent of total offences, approximately 5,417 cases, failed to progress due to evidential issues, even in cases where the suspect was identified and the victim remained willing to support the investigation. This points to a different problem: not victim engagement but investigative capability.

Digital forensics, platform preservation requests, metadata analysis and online evidence recovery are specialist skill, not capabilities that can be assumed across every frontline investigation

Is policing equipped to investigate these offences?

Refuge is calling for mandatory training for all officers on technology-facilitated abuse — not optional, not force-by-force, but mandatory and consistent across England and Wales. The government’s Violence Against Women and Girls Strategy and the recent police reform white paper both indicate an intention to improve training, but Refuge’s position is unambiguous: good intentions and guidance documents are not enough. The training must be mandatory, trauma-informed, culturally competent, and must reflect the reality that intimate image abuse is frequently not a standalone offence but part of a wider pattern of domestic abuse and coercive control.

This framing matters for how officers approach these cases from the first point of contact. A survivor reporting that her ex-partner has threatened to share intimate images of her is not reporting a niche digital crime. She is, in the majority of cases, reporting an act of coercive control perpetrated by someone she knows, someone who may have access to her home, her children, her finances and her wider life. The investigative and safeguarding response needs to reflect that reality.

The figures from Refuge’s technology-facilitated and economic abuse service underline the scale of what is coming down the line. Referrals to that service rose by more than 62 percent in 2025 compared to 2024, with the final quarter of 2025 the highest single quarter on record. Whatever forces are currently resourced and trained to handle, demand is growing significantly faster.

The Refuge data includes AI-generated images and videos within its definition of intimate image abuse, and this is increasingly where the investigative frontier lies. Deepfake technology, which allows a perpetrator to superimpose a victim’s face onto sexual content without her knowledge or consent, has moved in under a decade from a niche technical capability to a readily accessible consumer tool.

As Professor Clare McGlynn, one of the UK’s leading academic authorities on image-based sexual abuse, has documented extensively, the gap between what the law provides and what is actually being prosecuted is substantial. The intimate image abuse charging figures from Refuge are one of the clearest illustrations of that gap in operational terms.

What needs to happen

The Refuge data does not point to a single systemic failure. It points to several — in training, in investigative capability, in victim support, in data recording, and in the accountability mechanisms that would allow forces, HMICFRS and the Home Office to track whether improvement is actually happening.

On data recording alone, the fact that only 11 of 27 responding forces could provide a breakdown of the threat to share offence — a specific statutory provision that has been in force for five years — raises questions about whether forces are recording these offences consistently, whether the recording guidance is clear, and whether anyone is checking.

For officers on the frontline, the practical message is straightforward. Intimate image abuse is a serious, harmful and increasingly prevalent offence. It is not a relationship dispute. It is not a civil matter. It is not something to be redirected to a helpline. It is a crime, frequently embedded in a pattern of coercive and controlling behaviour, that requires an investigative response proportionate to the harm it causes.

The law exists. The question is whether policing is using it.

Data source: Refuge FOI requests to 43 police forces in England and Wales, responses received from 27 forces (25 providing outcome breakdowns), covering July 2021 to February 2026.