Debate contribution: Artificial Intelligence and Online Violence Against Women and Girls

Debate contribution by Mrs Naomi Long on .

MLA
Mrs Naomi Long
Party at the time
Alliance Party
Date
26 January 2026
Type
Debate contribution (made as a minister)
Business
Private Members' Business
Heading
Artificial Intelligence and Online Violence Against Women and Girls
Assembly identifier
Official Report component 5341483

Read the official Assembly record

Text

Thank you, Mr Deputy Speaker. I am very grateful to Cara and her colleagues for bringing this important motion before the House today, and specifically Cara and Diane, who, again, very powerfully shared their personal experience of the issue. Recent political, media and public focus on the issue of the use of X's AI chatbot, Grok, has brought into sharp focus how advances in technologies that may bring benefits to society are also used by those with malign intent to perpetrate further sexual abuse and exploitation against women and girls. Members will know that I am fully committed to strengthening the law to try to stop all abuse and violence against women and girls, and against men where that happens, although Members have rightly identified that the vast majority of victims of this abuse are women and girls. In the previous Assembly mandate, I introduced a new domestic abuse offence that captured controlling and coercive behaviour and psychological, emotional or financial abuse. It also covers online abuse, where abusers exploit the use of technology, social media and other online means to control or coerce their victim. I introduced new offences of stalking, upskirting, downblousing, cyberflashing and four new offences to tackle adults who pretend to be a child with a view to sexual grooming, which bolstered the child grooming offences. I created a stand-alone offence of non-fatal strangulation to provide greater protection against behaviour that is often used by abusers to exert control and create fear, serving as a precursor to further violence, and which is closely linked to domestic homicide. I enhanced protections around the offences of the abuse of a position of trust and the disclosing of private sexual images, which is more commonly referred to as revenge porn. Those offences have added real, tangible and valuable protections against online and offline abuse. The extension of the provisions in the UK Crime and Policing Bill to Northern Ireland will further strengthen those protections. The first of two intended legislative consent motions was agreed by the Assembly on 23 June 2025, and the second motion is scheduled for debate here on 2 February. They include measures to amend the current offence of the possession of a paedophile manual to bring AI-generated child sex abuse images within the scope of the offence definition, and to criminalise pornography that depicts strangulation and suffocation; and provisions for a new offence of child sexual image abuse generators to criminalise those who make, adapt, possess, supply or offer to supply anything that is made or adapted to create, or facilitate the creation of, child sexual abuse images.

Further new reserved offences that will be introduced by the Bill to extend across the UK include the creation of a new offence of online facilitation of child sexual exploitation and abuse and a new power to scan for child sexual abuse images at the UK border, meaning that phones could be taken at the border, when people are entering the country, and scanned for those images. Members will agree that, while that is a welcome and significant framework of protective measures, its necessity illustrates the level and nature of the harms, the level of abuse that is targeted at women and children and the dynamic and rapid changes that we have seen in those over recent years. Sadly, those with malign intent will always find new ways to continue the abuse. Women and girls now have to face the growing threat of having their image sexually manipulated online. I am deeply concerned about the exponential rise and the proliferation of online abuse in the form of non-consensual sexually explicit deepfake images. Unfortunately, AI tools have made it significantly easier to create such abusive images. That is an abhorrent form of sexual abuse and exploitation that is targeted overwhelmingly at women and girls. Many Members noted that research shows that deepfake pornography makes up 98% of all deepfake videos online, and that 99% of the individuals who are targeted in those deepfake videos are women. Everyone has a right to the protection of their image, their autonomy, their privacy and their dignity. Those images and the people who create and share them seek to take those rights away from their victims and to humiliate, harm and abuse them. The harm that is caused by the creation and sharing of such images is incalculable. Victims report feeling embarrassed, violated and unsafe as well as suffering from anxiety, post-traumatic stress disorder and suicidal ideation because of their experiences. People assume that so-called non-contact sexual offences are, in some way, less harmful or less real than contact sexual offences. Speak to any victim, and they will tell you otherwise. Knowing that a prospective employer who googles your name may find a deepfake image before they find out any of your qualifications or skills for a job is a profoundly disturbing experience for any person to go through, and many people do. In 2023, it was reported that more deepfake videos were viewed in that year than in all the preceding years put together. On the top deepfake sites, videos had been viewed more than 4·2 billion times. The psychological impact on individuals of knowing that those images are out there is severe. The proliferation of those images online has a wider societal impact. It perpetuates, promotes and normalises misogyny and the sexualisation of women. Such images dehumanise and objectify us. I believe that they also desensitise people, particularly men and boys, to the harm that such objectification can cause in real life. I will turn, briefly, to the impact on boys. It is extremely worrying that there is an increasing number of reports worldwide, including in Northern Ireland, of schoolchildren creating sexually explicit deepfake images of their classmates and other peers. That vile behaviour is already an offence under the Protection of Children (Northern Ireland) Order 1978 and is punishable on summary conviction by up to six months' imprisonment or on conviction on indictment by up to 10 years' imprisonment. It is vital, therefore, that boys know that creating such images can have serious consequences for them, lasting the rest of their lives. There is currently no legislation governing sexually explicit deepfake images of adults, but I am determined to address that gap and the threat and harm that such images cause. I will introduce a new deepfake offence by way of an amendment. It will be tabled at Consideration Stage of the Justice Bill, which is going through the Assembly. The timings will, ultimately, be dictated by the Assembly and the Committee, but the aim is to complete the amendment to introduce that offence in March. New offences to criminalise the creating —.

It is because of drafting pressures in the Office of the Legislative Counsel and because — I will get on to this later in my speech — we had to look at the outcome of the consultation. That is why. It is easy to eliminate the issue of vagrancy by repealing two pieces of legislation: that is a simple delete, following the consultation that had already been done in the previous mandate. This is a new consultation that had to be done at pace. The two are not comparable.

New offences to criminalise the creation of sexually explicit deepfake images of adults, the request for their creation, their sharing and the threat to share them will be included. The draft provision will be shaped by responses to the recent public consultation on legislative proposals. My intention is for every aspect of the despicable behaviour to be captured. That, of course, is the fundamental difference between the draft provision and other amendments: they were developed in previous mandates and are simply being introduced in this one. If enacted, the provision would bring Northern Ireland into line with England and Wales in providing for offences of creation and sharing and, in fact, would provide a more comprehensive package of measures than would have been possible via the LCM route. Members referred to retrospectivity. Article 7 of the European Convention on Human Rights states that no one shall be held guilty of a criminal offence for an act that was not a crime at the time that it was committed. In introducing criminal offences, therefore, it would be highly unusual and potentially in breach of article 7 to make them retrospective, which is why there is a degree of urgency in delivering the provision now. Unfortunately, the availability of AI tools or nudification apps has made the creation of the images all too easy. The UK Government recently announced their plans to ban such tools where their sole purpose is to use generative AI to turn images of real people into fake nude, sexually explicit pictures and videos. The proposed legislation, which will be tabled next month at Lords Report Stage as an amendment to the Crime and Policing Bill, would allow the police to target the firms and individuals who design and supply those tools. Regrettably, the plan to table an amendment to an existing amendment to the Bill means that it cannot be extended to Northern Ireland despite our best efforts. However, I very much welcome the recent announcement by my colleague Kate Nicholl of her plans to provide for similar legislation to make it illegal in Northern Ireland to create AI tools that enable non-consensual, sexually explicit AI-generated images. I hope that it will be possible to get consent from all parts of the Chamber to ensure that we keep pace with developments elsewhere. More widely, I will continue to do everything possible to keep women and children safe from harm and to protect them from all forms of online abuse and exploitation. Social media and online activity are increasingly central to our lives, but they also present serious risks. As technology evolves, so do the tactics of those who seek to cause harm. We cannot stand still when the creation, manipulation and sharing of sexually explicit deepfake images and child sexual abuse material using AI underscore the need to remain constantly vigilant. Our devolved powers are limited, given that telecoms is a reserved matter, so we need to work with Westminster to deliver a comprehensive response in this space. In that context, I highlight the UK-wide protection afforded by the Online Safety Act 2023. That places a duty, regulated by Ofcom, on social media companies to take steps to protect users from illegal content. If an organisation is considered to have broken the law, Ofcom can impose a penalty of a fine of up to 10% of that organisation's worldwide revenue, or £18 million, whichever is the greater. Where there is a failure to comply, Ofcom can seek a court order to force internet service providers to block access to the site in the UK altogether. Members will, no doubt, be aware of recent announcements by the Secretary of State for Science, Innovation and Technology on the UK Government's intention to issue a three-month consultation on further measures to keep children safe online, including, potentially, restrictions on addictive features, a ban on social media access and better age checks. As I did on the development of the Online Safety Act, I will continue to liaise with UK Government Ministers on any developments in that area. It is important to recognise that a criminal justice response alone will not fully address all the harms encountered online. The worrying reports of children making sexually explicit deepfakes of their peers emphasise the need to take a holistic approach focused on prevention and education, as set out in the cross-cutting Executive online safety strategy that has been put in place to assist children and young people who participate in the online world and want to do so in a positive, safe and responsible way. If my proposals for the new offences are enacted, my Department will liaise with the Department of Education to raise awareness of their introduction and of the existing offences that criminalise sexually explicit deepfake images of children. That needs to be part of the curriculum not only to equip young girls who are growing up with the threat of being the subject of such explicit deepfake images but to educate young men on the risks of engaging in what they may think is banter but what is a seriously damaging behaviour that has long-term consequences for them. I am sure that Members will agree with and support me on the steps being taken to address the despicable behaviours associated with deepfake images — behaviours that are unfortunately becoming increasingly pervasive. It is unacceptable that we live in a society in which some men and boys still think that it is acceptable to sexualise and objectify women and girls in that way. We have to do all that we can to protect our community from the harm that those and other forms of online abuse cause and seek to address the wider issue of misogyny and violence against women and girls, which is, sadly, all too prevalent in Northern Ireland. We will do that only by working together, but no one should be under any illusion. Even if we pass all this legislation, there will still be issues in the online space. Just today, I met campaigners for Caoimhé's law, who wish to see an end to people taking and uploading roadside photographs at car accident sites. This is an opportunity for people to talk about what socially responsible use of social media is and about the digital access that we have. The idea that people would send up a drone to photograph dead bodies lying at the side of a road after a car accident is repugnant on every level, particularly in the knowledge that the families of those people may see the images before they have been notified of what has happened. That is what passes for normal in today's society, however. It is therefore not enough simply to close down deepfake images or to have Caoimhé's law. It is about starting to engage as a society about what kind of people we are and what kind of people we want to be and then starting to educate our young people about discernment when it comes to using what are powerful tools when they are used for the good but can be utterly destructive tools when they are not.

2,395 words. Text as published by the Northern Ireland Assembly.

Assessment

Not flagged

Jev’s structured answers met the documented threshold for “not flagged”.

  • Overall probability that the activity is being used to promote, exploit or escalate an identity-based controversy: 0.05 (flagging needs at least 0.70; at or below 0.30 is not flagged).
  • None of the five culture-war criteria was met (each was below 0.60).
  • Assessed as legitimate policy scrutiny (0.95).
  • No identity-based framing was identified (0.74).
  • Subject area: Sexism, misogyny, ageism and other equality groups.
  • The overall probability was at or below the not-flagging threshold.

This is an AI-generated assessment of the text, made against published criteria. It is not a finding of fact about the MLA or their intentions. The Assembly record itself is the fact.

Jev’s structured answers

Probability that each statement is true of this text (0 = no, 1 = yes). A criterion counts as met at 0.60 or above.
QuestionProbabilityMet?
Identity framing0.09No
Sectarian or communal antagonism0.05No
Cultural grievance politics0.02No
Dog-whistle messaging0.05No
Controversy escalation0.05No
Legitimate policy scrutiny0.95Yes
Overall judgement0.05At or below not-flagging threshold
Context and attribution (stance)
OptionProbability
No identity-based framing (chosen)0.74
Reports or questions it0.00
Criticises or condemns it0.24
Adopts it0.02
Subject area (top five)
SubjectProbability
Sexism, misogyny, ageism and other equality groups0.74
Gender, sexuality, abortion and related social issues0.26
No identity-related subject0.00
Flags, emblems, symbols, names and memorials0.00
Parades, bonfires and commemorations0.00

Assessed 21 October 2026 at 14:31 UTC by jev-1.13.0 (requested as typesafe/jev) under rubric 2026.10-r9, methodology 1.9.

Disagree with this assessment?

Classifications can be challenged. A person reads every challenge, and any change is recorded publicly above with its reason.

Challenge this classification