Non-Consensual Sexually Explicit Images and Videos (Offences) Bill [HL] Debate

Full Debate: Read Full Debate
Department: Ministry of Justice
Lord St John of Bletso Portrait Lord St John of Bletso (CB)
- View Speech - Hansard - -

My Lords, I join in congratulating the noble Baroness, Lady Owen, on introducing this much-needed Bill updating the voyeurism laws as well as the patchwork of laws in line with the huge advances in this digital and social media age. I have had a long-standing interest in the deepfake threat and crisis from the days when I was a member of the Select Committee on Artificial Intelligence in the UK, chaired by the noble Lord, Lord Clement- Jones, way back in 2017. In chapter 8 of our report, we addressed the necessary measures to mitigate the risks of AI, but few of us contemplated the dramatic impact that Al would have on our daily lives—but also the huge threats of the violation of privacy and the psychological impact on individuals, particularly women, who have been subjected to deepfake sexual abuse, causing horrendous anxiety and trauma.

What I feel is imperative is the need to raise far more public awareness of deepfakes and improve prevention. Legislation is not enough; we also need to promote the development and use of advanced technologies to detect and prevent deepfakes and encourage much more co-operation between tech companies as well as policymakers to proactively address misuse. A purely technical solution to identify and label deepfake content is insufficient.

Certainly, social media platforms, not just adult content sites, have been financially incentivised to allow high-traffic content, including harmful deepfakes, because it drives advertising revenue. In this regard, the Bill goes a long way to impose severe penalties for failing to adopt robust technology solutions. Just as banks provide support for victims of scams, so too should similar assistance be extended to individuals harmed by deepfake abuse.

Finally, the proliferation of deepfake sexual abuse highlights the failure of digital identity verification. Generative Al can easily bypass facial recognition and centralised government ID systems. A more robust approach is needed, focusing on device fingerprinting and behavioural fingerprinting to blacklist users creating harmful content. These techniques will ensure that, even if such individuals attempt to reappear under new accounts or identities, they will remain blocked from platforms.

I wholeheartedly support this Bill and entirely agree with the noble Baroness, Lady Owen, that the legislation needs to be future-proof.