Deepfakes and Consent: The Law Catches Up
A Queen Mary Legal Advice Centre blog post by law student Fatimah Shah, tracking how UK law closed the gap on deepfakes. Sharing a non-consensual intimate image was already an offence, but making one was not until the Data (Use and Access) Act 2025 came into force in February 2026. The piece also covers the shift that matters most: proving the subject did not consent, rather than proving the perpetrator meant to cause distress.

What changed
For years the UK position had an obvious hole in it. Sharing a non-consensual intimate image was an offence. Making one was not. Someone could generate a synthetic intimate image of you, keep it, and have committed no crime until they passed it on.
Shah's piece walks through the legislation that closed that gap.
- Online Safety Act 2023. Amended the Sexual Offences Act 2003 to criminalise sharing intimate images without consent, explicitly including images that only "appear to show" a person. That brought deepfakes inside the sharing offence.
- Data (Use and Access) Act 2025.* In force from **6 February 2026**. Inserted new offences into the Sexual Offences Act 2003 criminalising the *creation of purported intimate images of an adult without consent. This is the one that matters most, because it moves the line from distribution to production.
- Crime and Policing Bill 2025*, third reading *25 March 2026. Proposes further offences covering creation and dissemination, and possession of AI models built to generate child sexual abuse material.
For images of children the law was already there, through the Protection of Children Act 1978, the Criminal Justice Act 1988 and the Coroners and Justice Act 2009, all of which already covered computer-generated indecent images.
The scale of the problem
The figures Shah cites are the argument for why this took priority:
- Of 15,000* deepfake videos catalogued by Deeptrace in September 2019, *96% were pornographic.
- 99% of sexual deepfakes target women.
- By 8 January 2026*, Grok was receiving roughly *6,000 requests an hour to generate sexualised images of people.
Danielle Citron, the Boston University law professor whose work underpins much of this field, puts it plainly: "Deepfake technology is being weaponised against women."
Consent, not intent
The most important legal shift here is about what a prosecutor has to prove.
Older image-based abuse law required showing the perpetrator intended to cause distress. That is a high bar and a strange one, because it makes the victim's protection depend on the state of mind of the person who harmed them. Someone who did it for money, for status, or for a laugh could fall outside it.
Georgia Harrison's campaign, following her own case in 2023, drove the reform that removed that requirement. Her line on it: "The reforms to the law that has been passed today are going down in history as a turning point for generations to come."
The #StopImageBasedAbuse coalition wants the same logic applied throughout: the offence should turn on whether the person consented, not on what the perpetrator was hoping to achieve.
What is still missing
Shah is measured about it. Campaigners regard these measures as real progress and as the beginning rather than the end. Enforcement is the obvious open question. An offence on the statute book requires police capacity, prosecutorial appetite, and platforms that respond faster than the images spread.
Why it matters
This sits under creative rights because it is the same principle arriving from a different direction. Your face, your voice and your likeness are yours. A system that can synthesise any of them does not acquire the right to do so by being technically capable.
The interesting thing for anyone tracking AI policy is the sequencing. Rules governing synthetic media landed first, and landed relatively fast, in the area where the harm was most undeniable and most gendered. The arguments about training data, attribution and payment are still grinding through consultations. The law here moved because the harm was impossible to look away from.
Worth reading with its provenance in mind. This is a student blog post from the Queen Mary Legal Advice Centre, clearly written and well sourced, but a clear explainer rather than legal advice.
Key takeaways
- Creating a non-consensual intimate image, including a deepfake, became a criminal offence for adults in the UK on 6 February 2026 under the Data (Use and Access) Act 2025.
- Before that, sharing was illegal but making was not.
- The reform moved the test from the perpetrator's intent to the subject's consent, largely thanks to Georgia Harrison's campaign.
- Equivalent protection for images of children already existed under earlier legislation.
- Campaigners regard this as a start, with enforcement the open question.
This page covers image-based abuse. In the UK, the Revenge Porn Helpline supports adults affected by intimate image abuse, and Report Remove, run by the NSPCC and IWF, helps under-18s get images taken down.
· End of dispatch ·
About the correspondent

Dead Good Club
The House Curator brings together stories worth knowing about. Our editor selects and rounds up useful resources for the library, while summaries (usually AI-generated) offer a quick overview and helpful context. Every post links to the original article or video, so you can explore the full story for yourself. If you'd like to write for The Dead Good Club and see your work featured here, get in touch by email.




