Following an election manifesto promise, the new government is expected to bring  in  a law to  ban the creation of sexually explicit deep fakes.  It was not named in the King’s Speech, but is predicted to feature in  the Crime and Policing Bill, that addresses violence against women and girls. 

 As such, it will amend the criminal law and fall under the remit of the Home Office. It will almost certainly have to  be linked to the Online Safety Act, which already contains provisions to criminalise  the sharing of intimate images.  The new law would work in tandem with the existing provisions, and strengthen them. 

Deep fakes are a specific issue arising out of artificial intelligence (AI) technology  which can create hoax images, including 3-D and action video, that purport to show a real person. Sexually explicit deep fakes can be very distressing for the victim.

Under the Online Safety Act, cyberflashing and so-called revenge porn are criminal offences, as is the sharing of so-called deep fake porn.  Anyone sharing intimate images on social media or messaging platforms without consent, or with intent to cause alarm or distress, could fall foul of the law and end up in court. It is expected that the new proposal to criminalise "creation" of sexually explicit deep fakes would form a trio  that would strengthen powers to tackle a range of issues concerning the sharing of sexually explicit images. 

The two offences in the Online Safety Act 2023 are  S.187 addressing cyberflashing and S.188 which is the so-called revenge porn clause. Both of these offences are amendments to S. 66 of the 2003 Sexual Offences Act, and they act as a pair. They were both inserted into the Online Safety Act at  the House of Lords Report stage.

Under the S.187, the  sharing of photographs of anyone’s genitals is an offence, if the intention is to cause distress, alarm or humiliation or if the person sharing is wilfully reckless of that possibility. This has  already been tested in a court. Notably, a conviction was handed out by the Southend Crown Court on 19th March this year. The defendant pleaded guilty to  two counts of sending a photograph or film of genitals to cause alarm, distress, or humiliation. LINK   Cyberflashing  is not a trivial offence, as it may be a precursor to very grave offences such as rape.   

Under S.188, it is an offence to  share an intimate image of someone who does not consent to it being shared. It is also an offence to share an intimate image of someone with the intention of causing that person distress, alarm, or humiliation,   or threatening to share such an image knowing that the person will fear it  being disseminated.  The definition of  an intimate image under S.188 states: all or part of the person’s exposed genitals, buttocks or breasts which also includes, for example, where they are visible through wet clothing.

So how would this apply to deep fakes? It is already a criminal offence under S. 187 and S.188 to  share sexually explicit deep fake  images without consent, or with intent to cause alarm, distress or humiliation . It is also an offence to [paraphrasing] use a sexually explicit deep fake image to threaten or effectively blackmail, someone.  The Act does not mention the words deep fake but it does refer to whether an image is made or altered by computer graphics or in any other way, which appears to be a photograph or film. The expected new proposal would mean that not just sharing or using the image, but making it, would be an offence.   

Labour committed to the policy in its election manifesto which stated:  "Labour will ensure the safe development and use of AI models […] by banning the creation of sexually explicit deepfakes. "  Apparently, it had been considering a proposal  to go a little further and also ban the tech tools that are used to create of deep fakes. That idea was based on a  policy paper from a the  Labour Together  thinktank, although it’s not known if that will be followed through.   The  measures announced  follow an earlier, similar  initiative on deep fakes by the previous government.    

The measures in S.187 and S. 188  were amendments proposed by the former Culture Minister Maria Miller. They have had more attention to  detail in the drafting than the majority of the Online Safety Act. The measures are  clear and specific, both in terms of the policy aim and the manifestation in an online post. They do not form part of the S.10 measures  and if I am correct, would not be subject to proactive monitoring.  

---

If you cite this article, kindly acknowledge Dr Monica Horten as the author and provide a link back.

I provide independent advice on policy issues related to online content, Please get in touch via the contact page.