Legislate New Offences for Violence Against Women and Girls
Conservative · what the evidence says
An independent, source-checked look at Conservative’s policy “Legislate New Offences for Violence Against Women and Girls” — what it would actually do across the things that affect your life. Every claim below quotes the source behind it. How this works.
Personal liberty & free speech — Hurts
minor · moderate confidence
This policy creates new criminal offences and strengthens existing ones, which by definition expands the state's coercive reach over individuals' conduct. The liberty cost is real but narrow — the new offences target specific non-consensual acts rather than broad speech or protest.
The evidence
- The policy commits to legislating new offences for spiking, creation of sexualised deepfake images, and taking intimate images without consent. — conservatives.com (manifesto) — “legislate to create new offences for spiking, the creation of sexualised deepfake images, and taking intimate images without consent”
- The policy also commits to strengthening existing laws against stalking, coercive behaviour, and revenge porn. — conservatives.com (manifesto) — “strengthening existing laws against stalking, coercive behaviour, and revenge porn”
- Perpetrators creating and sharing deepfakes could face up to two years in prison under related legislation. — gov.uk (media) — “Perpetrators creating and sharing deepfakes could face up to two years in prison”
- Advocates highlight an existing implementation gap, with many existing legal powers already underused, suggesting new offences may not fire at scale. — committees.parliament.uk (government) — “many existing measures are not being utilised in day-to-day practice, pointing to negligible use of some powers”
Biggest unknown: Whether courts or prosecutors interpret offences (especially 'sexualised deepfake images') broadly enough to capture lawful expression or satire, which would materially worsen the liberty impact.
Our reading: O10 scores the liberty effect of state coercion on individuals, regardless of whether the conduct targeted is harmful. By creating new criminal offences and strengthening existing ones, this policy extends the reach of state coercive power — individuals who would previously have faced no criminal liability now face prosecution and imprisonment. That is a textbook O10 worsening. The magnitude is minor because the offences are narrowly targeted at specific non-consensual acts (spiking, non-consensual image creation/sharing, stalking) rather than broad categories of speech, association or protest. There is no surveillance mandate, no ID scheme, and no compulsory licensing. The liberty cost falls almost entirely on would-be perpetrators of acts that already violate others' bodily autonomy and privacy — the population affected is narrow. A genuine residual risk is definitional overreach: 'sexualised deepfake images' could in principle be drawn broadly enough to capture satire or artistic content, but the policy text does not resolve this. The implementation-gap evidence (E33) also suggests new offences may not translate to materially more prosecutions, which would limit both the crime-reduction benefit (O5) and the additional liberty cost scored here. On balance: a real but minor, targeted expansion of criminal law — worsens O10 modestly.
Crime, justice & national security — Helps
minor · moderate confidence
Creating new criminal offences for spiking, deepfake images, and non-consensual intimate images closes genuine legal gaps in protecting women from serious harms, which should improve safety and access to justice. However, a persistent implementation gap — where existing powers already go largely unused — means new laws alone are unlikely to produce large real-world reductions in crime.
The evidence
- The policy commits to legislating new offences for spiking, sexualised deepfakes, and non-consensual intimate images, and to strengthening existing laws on stalking, coercive behaviour, and revenge porn. — conservatives.com (manifesto) — “legislate to create new offences for spiking, the creation of sexualised deepfake images, and taking intimate images without consent, alongside strengthening existing laws against stalking, coercive behaviour, and reveng…”
- Over one million VAWG-related crimes were recorded by police in 2022/23, around 3,000 offences per day. — commonslibrary.parliament.uk (government) — “Over one million VAWG-related crimes were recorded by the police in 2022/23, equating to about 3,000 offences per day”
- Recorded VAWG-related crime increased by 37% between 2018 and 2023. — commonslibrary.parliament.uk (government) — “Recorded VAWG-related crime increased by 37% between 2018 and 2023, with one in every twelve women being a victim per year”
- An estimated 98% of all deepfake content online is non-consensual and pornographic, and 99% of those depicted are women. — stimson.org (media) — “98% of all deepfake content online is non-consensual and pornographic, and 99% of those depicted are women”
- Existing comparable powers are barely used: Domestic Violence Protection Orders are used in only 1% of domestic abuse cases, and coercive and controlling behaviour is charged in only 1.75% of prosecutions. — committees.parliament.uk (government) — “Domestic Violence Protection Orders used in only 1% of domestic abuse cases, coercive and controlling behaviour charged in 1.75% of prosecutions”
- Victims of rape and domestic abuse withdraw from the criminal justice process at very high rates — 69% and 59% respectively. — victimscommissioner.org.uk (media) — “Victims of rape and domestic abuse withdraw from the criminal justice process at disproportionately high rates (69% and 59% respectively”
- The Centre for Women's Justice identifies a significant implementation gap between legal measures and their on-the-ground effectiveness, arguing many existing measures are not used in day-to-day practice. — committees.parliament.uk (government) — “significant "implementation gap" between legal measures and their effectiveness on the ground”
- The NAO questioned the effectiveness of past VAWG strategies, citing lack of buy-in, failure to learn from past strategies, and limited progress on prevention. — commonslibrary.parliament.uk (government) — “National Audit Office (NAO) in January 2025 questioned the effectiveness of past VAWG strategies, citing a lack of "buy-in" from government departments, failure to learn from past strategies, no consistent definition of …”
Biggest unknown: Whether police forces will actually use the new offences in practice, given documented evidence that existing comparable powers (e.g. coercive control charges, DVPOs) are currently used in only a tiny fraction of eligible cases.
Our reading: The policy addresses real and significant harms. VAWG affects enormous numbers of people — over a million recorded crimes annually, rising sharply — and areas like deepfake intimate imagery represent a rapidly growing harm with a clear legal gap (E26, E27). Creating specific criminal offences for these behaviours sends a deterrence signal, enables prosecution where it was previously difficult or impossible, and affirms the justice system's protective function for victims. On O5, this is a genuine improvement: where conduct was previously outside the criminal law or only partially covered, victims gain a route to justice and perpetrators face legal sanction. However, the magnitude must be kept modest. The central problem in VAWG enforcement is not primarily the absence of legislation but the failure to use existing powers — coercive control is charged in under 2% of eligible prosecutions, DVPOs in 1% of cases (E33). High victim withdrawal rates (E31) and the NAO's finding of weak inter-departmental buy-in and no consistent strategy (E37) suggest that adding new offences without resolving these structural problems will not produce large real-world crime reductions. The CWJ's implementation gap analysis (E32) directly applies: new law on the statute book does not automatically translate into practice. Absent the policy, conduct such as sexualised deepfake creation currently sits in a partial legal grey area; the policy's marginal addition is real but limited to closing that gap rather than transforming enforcement culture. The direction is 'improves' because closing genuine legal gaps is a concrete protective benefit — victims gain legal standing they lacked before — but magnitude is 'minor' because the evidence strongly suggests enforcement culture and resourcing are the binding constraints, not the statute book.
Equal treatment & democratic rights — Helps
minor · moderate confidence
Creating new criminal offences for spiking, deepfake images, and non-consensual intimate images strengthens formal legal protections for women, who face these harms far more than men. However, evidence shows that new laws in this area often sit unused without the resourcing, training, and culture change needed to make them work in practice.
The evidence
- The policy commits to legislating new offences for spiking, sexualised deepfake images, and non-consensual intimate images, and to strengthening existing laws on stalking, coercive behaviour, and revenge porn. — conservatives.com (manifesto) — “legislate to create new offences for spiking, the creation of sexualised deepfake images, and taking intimate images without consent, alongside strengthening existing laws against stalking, coercive behaviour, and reveng…”
- Women are significantly more likely than men to experience domestic abuse, stalking, sexual assault, and harassment, establishing the gendered basis of these harms. — commonslibrary.parliament.uk (government) — “women were more likely to have experienced violent crime, domestic abuse (6.6%), stalking (4.0%), sexual assault (3.4%), and harassment (11.2%) compared with men (3.0%, 2.3%, 0.8%, and 6.6%, respectively)”
- Deepfake harms are overwhelmingly concentrated on women, with 99% of those depicted being female. — stimson.org (media) — “98% of all deepfake content online is non-consensual and pornographic, and 99% of those depicted are women”
- Existing VAWG protective powers are drastically underused — Domestic Violence Protection Orders used in only 1% of cases, coercive control charged in 1.75% of prosecutions. — committees.parliament.uk (government) — “Domestic Violence Protection Orders used in only 1% of domestic abuse cases, coercive and controlling behaviour charged in 1.75% of prosecutions”
- The Centre for Women's Justice identifies a significant implementation gap between legal measures and real-world effectiveness, attributing this to lack of resourcing, training, and culture change in policing. — committees.parliament.uk (government) — “The Centre for Women's Justice (CWJ) highlights a significant "implementation gap" between legal measures and their effectiveness on the ground”
- The NAO questioned the effectiveness of past VAWG strategies, citing lack of government buy-in, failure to learn from previous strategies, and limited prevention progress. — commonslibrary.parliament.uk (government) — “The National Audit Office (NAO) in January 2025 questioned the effectiveness of past VAWG strategies, citing a lack of "buy-in" from government departments, failure to learn from past strategies, no consistent definition…”
- Victims of rape and domestic abuse withdraw from the justice process at very high rates (69% and 59% respectively), limiting the practical reach of any new offences. — victimscommissioner.org.uk (media) — “Victims of rape and domestic abuse withdraw from the criminal justice process at disproportionately high rates (69% and 59% respectively, based on London-specific data, but acknowledged as a national issue)”
Biggest unknown: Whether new offences translate to real prosecutions, given that existing VAWG powers are already severely underused — coercive control, for instance, is charged in only 1.75% of relevant cases.
Our reading: This policy directly advances O9 by extending formal legal protection against harms that fall overwhelmingly on women. Criminalising deepfake creation and non-consensual intimate images, alongside stronger stalking and coercive control laws, expands the anti-discrimination and equal-treatment framework in law. The gendered baseline is stark: women experience domestic abuse, stalking, sexual assault, and harassment at roughly double the male rate, and 99% of deepfake victims are women — so new offences here have a clear and targeted equal-treatment rationale. However, the direction is constrained to 'minor' by compelling implementation evidence. The historical record on VAWG legislation shows that legal instruments routinely fail to translate into prosecution practice: coercive control is charged in just 1.75% of cases despite existing law, and protection orders are used in 1% of eligible domestic abuse cases. The NAO found the wider VAWG strategy lacked departmental buy-in and failed to learn from past failures. High victim withdrawal rates (69% for rape, 59% for domestic abuse) further erode the realistic population-scale effect of any new offence. Absent the policy, gaps in the legal framework remain — particularly for deepfake creation (which, on the evidence, has been partly addressed through subsequent legislation by a different administration). The genuine marginal addition of this policy, compared to the baseline of legislative activity already underway, is therefore uncertain. What the policy clearly does is extend the formal equal-treatment architecture, which is real and non-trivial; what it cannot guarantee — and the evidence strongly suggests will not automatically follow — is meaningful real-world enforcement at scale. The verdict is 'improves/minor': a genuine advance in legal status, partially offset by the persistent implementation gap that is the central caveat of all VAWG law reform in England and Wales.