Harassment, Harmful Communications and Related Offences (Amendment) (No. 3) Bill 2026: Second Stage
The Bill would amend Coco’s Law to address non-consensual AI-generated intimate images, including their creation and commissioning, and provide routes for victim identification and action against platforms. Senators broadly supported it, citing the serious harm to victims and the need to keep pace with technology; Senator Keogan backed its aim but warned of drafting, free-expression and enforcement concerns. The Minister said the Government would not oppose Second Stage, while noting legal and procedural questions and related EU measures under way. Committee Stage was agreed for the following Tuesday.
No. 96 of 2026 ›
The combined opening speeches of the proposer and seconder shall not exceed 16 minutes. All other Senators will have six minutes and the Minister will have 15 minutes. The proposer will have five minutes in which to reply. Cuirim fáilte roimh an Aire.
Comment on this
I move: "That the Bill be now read a Second Time."
The Civil Engagement Group is very proud to bring forward yet another piece of legislation that will improve people's lives on social media. I am proud to stand in front of the Minister this evening with a Bill to amend Coco's Law. We have worked on this legislation for the past 18 months. I thank Corey O'Hara for all his hard work. I also thank the solicitors from the Oireachtas for all of their hard work on this very important legislation.
There are three really important strands to the Bill. They are: reduce harm and protect victims; empower victims and create a clear pathway to justice; and empower the courts when it comes to harmful technology. For us, this legislation is straightforward. It mostly protects women and children, although we recognise that men can also be victims of online abuse. How AI works in Ireland and beyond is that people can up inappropriate images of children, men, women or anybody that they want to abuse on social media. Before all of the scandals came out about Grok, we were working on this important legislation. It follows on from the child trafficking legislation that we brought forward to the House, which I am proud that the Minister is bringing before the Dáil.
I am not going to talk for much longer. In my opinion, this is very clever legislation that will amend Coco's Law. It is not about reinventing the wheel; it is about using what we have to implement a better justice system for all people who are impacted by inappropriate images on social media.
Comment on this
I commend Senator Flynn and her office on leading on this Bill. They have previously been very active on the proper and appropriate labelling of child sexual abuse material.
The potential negative impacts of the commercial models within some social media platforms and providers is something Senator Ruane and I highlighted when the data protection legislation was going through back in 2018. Section 30 of that legislation, which, unfortunately, has never been commenced, was specifically designed to tackle the commercial targeting of children through micro-profiling. There are also other measures that we highlighted over the period.
When the scandal relating to Grok arose about a year ago, the initial instinct of the Minister of State, Deputy Smyth, that it was a criminal matter was correct. We were then told that the way Coco's Law was being interpreted in the context of relating to a person who distributes, publishes or threatens to distribute material and that this did not allow for the pursuit of platforms or publishers. Grok is only one of the instances of the kind of deepfakes and nudification apps we have seen. It is my view that if it is a situation whereby, in response to a prompt, Grok produces a nudified image of a woman or child in the public space, Grok is in fact distributing that image. Our laws should have been used in a stronger way at that time and applied more directly in terms of this being considered a criminal offence. In any event, the legislation that is being put forward today is designed to ensure that any perceived loophole will be being addressed.
The Bill before us addresses a number of things in very particular ways. It includes a specific definition of what constitutes a deepfake intimate image in order to ensure that the current prohibitions in relation to an intimate image remain as they are and that there is no perceived dilution or even a perception that, previously, what was considered an intimate image did not include a deepfake intimate image. That categorisation remains intact, but it includes, for useful purposes, a specific categorisation of deepfake intimate image as an additional definition that is useful in terms of the application of the law. The Bill contains certain provisions in respect of the anonymity of victims. Crucially, it also contains measures in relation to the platforms, providers and companies that offer these services. The Bill makes it very clear that actions can, should and must be taken. If we are serious about tackling the abusive use of deepfake images of people that are used without their consent - because consent matters - then we need to ensure that we tackle the companies and not solely the individuals who may, for example, produce images by means of a prompt.
It is important that the penalties in this legislation match those in Coco's Law. However, the Bill also introduces a new category and a new set of responses that Coimisiún na Meán would be empowered to employ in respect of the platforms, publishers and companies that are in many cases facilitating and profiting from the production or the threat of production of these kinds of images.
When debates of this type take place, there is often a tendency to point towards the EU and the legislation it has brought forward. In that context, I want to touch on a couple of the reasons that what we have in terms of EU legislation in respect of tackling the companies to which I refer falls short. The AI Act is framed very much in terms of a product safety rule book rather than the criminal law. Much has been said about Article 50, which came into force in August, in terms of the requirement for the labelling of deepfake images, but it is not an issue for people to know that they are consuming deepfake images. This is also not solely a matter of disinformation, even though we know that deepfake images have been used in terms of spreading political disinformation and interfering in elections.
None of this addresses the harm such uses of AI do to individuals who may have been nudified, intimidated or personally targeted, regardless of whether the content is labelled as a deepfake. None of that is addressed by the simple matter of labelling. Even when we go to the political space, the political damage done to public figures through nudification and the creation of abusive images, of which multiple examples could be given, is not solely a matter of disinformation. What is done to those figures is matter of intimidation. It is to intimidate people to prevent them from participating in the public space. That is alongside its use for direct abuse and objectification, particularly of women and children.
There are loopholes in terms of general-purpose AI models under the current AI Act. I have to flag as an extremely serious concern the discussion that is happening right now in relation to the digital omnibus. One of the crucial decisions that will be discussed and made during those negotiations is the decision as to whether access will be given to personal data, which would include biometric personal data. There is heavy lobbying to provide access to personal data for use in the training of AI models. Looking at the irresponsibility with which many of the major companies proposing AI initiatives and many of the largest actors in this sector have acted and the idea of giving access to personal data for the purposes of training AI, which has been shown to be being deployed in a "move fast and break things" model rather than through responsible action being taken, while I know this is not directly within the Minister's remit, I would urge the Irish Government to take appropriate action to make sure that Ireland does not support or facilitate a watering down of personal data protection or the removal of consent requirements for the use of personal data simply to facilitate the commercial development and training of AI models. That is crucial. The very fact that this is being lobbied for is, in itself, a concern.
I look across to the United States where, just yesterday, a very strong message was sent not simply in terms of the electoral turnaround we have focused on, but in measures specifically relating to the regulation of data. In numerous states, numerous propositions got overpowering votes of support. In local communities, 70% or 80% of voters voted to show their concern at what they felt was an under-regulation of the digital and data areas. That is important context because the regulatory pieces that are happening here in Ireland very much relate to what is happening in the United States. Within the European context, we are the core regulatory centre. We are watched closely for that reason. We also have a great responsibility for that reason.
I will not go into the specific issues with different AI companies but one reason we need clear regulation directed at platforms and companies is that, without that regulation, we cannot trust that they will act responsibly. They themselves are asking for regulation. It is crucial that they do not simply set it themselves. Anthropic recently proposed a draft Bill that suggested that regulators should step in on AI models that had the potential to kill more than 50 people or to do more than $1 billion worth of damage. If that is the bar the industry is setting itself, we clearly need to set a different bar, one that is in line with the promise made to the public through GDPR and in line with the public's expectation that action will be taken, an expectation we saw in the response to the great distress that Grok's nudification and that wave of apps caused.
This is an opportunity. I hope that the Government will support this legislation. It is very clear. It takes existing legislation and ensures it is applicable in the context of deepfake and intimate images. It goes further, ensuring that the platforms and companies are responsible. Some of those who have provided feedback have argued that the penalties should be even higher. The criminal penalties in this section apply to the individuals who are acting. The penalties attaching to companies relate to their not blocking tools to ensure that there is not continued access to them where they can reasonably be said to be failing to act on the potential use of their services for these purposes or to be taking inadequate action or steps, which is a very reasonable bar. I really cannot see how the Minister can possibly disagree with this. It is very reasonable.
I commend the Office of Parliamentary Legal Advisers, OPLA, which has worked with Senator Flynn's office, and Corey Whyte, who has worked with her on the drafting of the Bill. I see this Bill as the bare minimum. I personally think that we need to step up in a wide number of areas relating to the regulation of the use of AI and to strengthen such regulation. This is an essential and basic step in addressing this issue, which hits people at the most personal level, can create lifetime repercussions and distress for the individuals affected, and does damage to society by intimidating people away from political and public participation.
I urge the Government, Government Senators and others to take action and to support the Bill. I look forward to the response. I hope that the Bill will not only be supported, but that it will be moved through rapidly because we are seeing an absolute explosion. In the main, it is not the major companies we know of but I have seen reports of smaller apps that appear to do one thing but present you with options to generate deepfake images once you go into the options menu. We need to put a brake on the tidal wave. Acting as soon as the Grok issue arose would have been better but acting now is the next best thing we can do.
Comment on this
The Minister, Deputy O'Callaghan, is very welcome to the House this afternoon. The Fianna Fáil group very much welcomes the objectives of the Bill and will not be opposing it. We are committed to making the online world a safer place for everyone and the measures proposed here are broadly aligned with our zero-tolerance approach to domestic, sexual and gender-based violence. This Bill seeks to amend the Harassment, Harmful Communications and Related Offences Act 2020, commonly known as Coco's Law, to explicitly recognise deepfake intimate images of a person generated or manipulated using artificial intelligence. This is an important issue because advances in artificial intelligence have created new opportunities for technology to be misused, with devastating consequences for individuals and families.
We have seen growing public concern following the introduction of AI-powered image editing features, particularly the Grok AI feature on the X platform, which Senators mentioned here in December last year. That technology was used on a significant scale to generate harmful and illegal material. This included the manipulation and nudification of images of women and children. What makes this particularly concerning is the speed and scale at which such images can be generated, shared and distributed. A person can find themselves the subject of an intimate or sexually explicit image without ever having consented to it and without that image having existed in reality. The consequences are very real, however. Victims can experience humiliation, distress, reputational damage, anxiety and a profound loss of privacy. Once an image has been circulated online, it can be extremely difficult to remove it and prevent further distribution.
We must also recognise that women and children are disproportionately targeted by this type of abuse. It is not simply a technological issue. It is one of dignity, personal safety, consent and basic human rights.
It is important to clarify what the existing law provides. Coco's Law makes it an offence to distribute or publish intimate images without consent, including when there is an intention to cause harm or distress. However, the legislation does not generally criminalise the mere creation of an image, including one generated with artificial intelligence. This distinction is significant. While the non-consensual sharing of intimate images is already illegal, a gap in our legislation remains in relation to the creation of such images involving adults. The Attorney General's advice in January 2026 confirmed this position. It also highlighted that the existing definition of an intimate image was sufficiently broad to cover digitally generated or manipulated images, meaning that the issue is not necessarily the defamation, but the need to address the act of creation. I also emphasise that the law already provides protections for children under the Child Trafficking and Pornography Act 1998. The creation of child sexual abuse material, CSAM, is an offence, regardless of whether artificial intelligence, AI, has been used to generate or alter an image and regardless of whether it depicts an actual child. Nevertheless, we must continue to strengthen our legislative response to keep pace with rapidly developing technologies.
There are important developments at European level that must also be considered. The EU violence against women directive provides for the criminalisation of creating deepfake intimate images and subsequently distributing them. Ireland is required to transpose the directive into national law by 14 June 2027 and I understand that work on the necessary legislative reforms is already under way. In addition, the EU AI Act introduced measures aimed at the development and use of particularly harmful AI systems. Recent changes provide for a prohibition on AI systems generating CSAM or creating sexually explicit images, videos or audio depicting identifiable individuals without their consent, unless the necessary technical safeguards are in place. Companies have until 2 December this year to bring their systems into line with these requirements.
Ireland has always been a strong advocate at European level for these protections and it is important that we continue to play an active role in ensuring technology companies are held accountable. While there are legal and practical considerations around how an offence relating to the creation of deepfake images can be effectively policed and protected, these should not prevent us from examining how our laws can be strengthened. We must strike the right balance between protecting individuals, ensuring effective enforcement and keeping our legislation up to date with technological developments.
There are a number of important principles we must underpin in our response. First, consent must remain central. No one should have their image manipulated into sexually explicit or intimate material without their permission. Second, technology companies must take responsibility for the platforms and tools they provide. They cannot simply stand back when their products are being used to facilitate abuse and exploitation. Third, victims must know that there is somewhere to turn when they are in difficulty. Anyone concerned about intimate images being shared online should report the matter immediately to An Garda Síochána and-or hotline.ie. Ultimately, we must ensure that the law protects people both online and offline. The fact that abuse takes place through AI or a digital platform does not make it any less serious and does not lessen its impact on those targeted.
We cannot allow technological advances to outpace the provisions available to citizens. I welcome this discussion. I acknowledge the work done by Senators opposite in this space and I look forward to this legislation passing through the Houses.
Comment on this
I welcome the Minister to the House. I acknowledge Senators Flynn, Higgins, Ruane and Black for bringing this legislation before the House. We can all agree that AI has the potential to do enormous good, but as with any advance in technology, there are always people who will find ways to do harm and damage with it.
What we are discussing is particularly worrying. That someone can take a perfectly innocent picture of a young boy or girl, a teacher, a colleague or anyone and within seconds turn it into a sexually explicit image should be of real concern to every one of us, and it is. While we might know the image is fake, the damage it can do to someone's reputation, confidence and, most important, mental health is real. Senator Higgins referred to that already. We made significant progress with Coco's Law, but the reality is that technology is moving at such a pace that legislation is struggling to keep abreast of it. That is why I welcome the intention behind the Bill.
It is important that we recognise the work that has already been done. With Senator Ruane, I sit on the Oireachtas Committee on Artificial Intelligence. The Minister of State, Deputy Niamh Smyth, has been actively working at European level to strengthen the EU AI Act, including on the prohibition of the creation of these images. The committee has been examining the issue and the Government is progressing legislation to implement the EU violence against women directive by June next year. Work is going on at the moment, but the important thing is that we bring all of this work together rather than end up with many different Bills that overlap or potentially create confusion.
As Leas-Chathaoirleach of the Committee on Education and Youth, I regularly hear about the challenges that young people, parents, school management and even teachers are facing. Are we doing enough to educate young people about the consequences of creating or sharing these images? Something that was suggested in a policy paper launched by Deputy Geoghegan yesterday was the further expansion of digital literacy certification and more work being done on educating young people in schools, but also the working population and the general population, to make an e-learning element available to everyone. Do our schools have clear guidance on what they should do when something like this happens with deepfakes? Are we supporting young people who find themselves at the centre of this? Prevention and education have to be every bit as important as prosecution in such situations.
I will highlight enforcement in this area. We can pass all the legislation we want in this House, but unless An Garda Síochána and the relevant regulatory authorities have the expertise, resources and powers to deal with the offences, we will not achieve what we set out to do. I would welcome hearing the Minister's views on how this Bill would work alongside the legislation already being progressed at European and national levels and whether there are opportunities to strengthen the protections for victims without duplicating the work already under way.
I welcome the Bill and certainly support the Senators and what they are trying to achieve, but this has to be about more than passing another Bill. It has to be about making sure that when people are targeted in this way, they know where to turn and where they can get the support they need and that those responsible will be held accountable.
Comment on this
I, too, commend colleagues in the Civil Engagement Group on bringing the Bill to the House. While I fully support the goal of stopping people using AI to harm private citizens, the Bill is slightly unworkable, full of legal issues and blind to how the courts operate.
Well-meaning sentiment is poorly served by bad law. We need practical rules that work. Let us start with the core of the Bill, which makes it a crime to generate a deepfake image under section 2A. To get a conviction that carries a heavy, seven-year prison sentence, prosecutors must prove beyond reasonable doubt that the creator had the explicit intent to cause harm and act recklessly.
This is where things become blurred. By using the highly subjective standard of recklessness, this Bill creates a massive loophole that cuts both ways. The whole point of biting political satire, comedy or caricature is to cause discomfort or shock to public figures. Because this text contains no clear protection for artistic expression or parody, it will create an immediate chilling effect. A satirist using modern AI tools to critique a politician could easily be targeted by a powerful individual claiming that the caricature caused him or her severe alarm or distress. We risk passing a law that powerful figures can weaponise to shut down legitimate dissent. Yes, in theory, many satirists would be exonerated but we have to acknowledge that at this moment in time, the process itself is a punishment.
Furthermore, this Bill creates deep confusion when mixed with existing law. We must remember that Coco's Law already covers images that purport to be intimate images of a person. Under current law, distributing and sharing these harmful fakes is already a severe criminal offence. The primary loophole this Bill attempts to close is the mere act of creating and keeping a file on a private device but by forcing prosecutors to prove a person's exact state of mind, we are handing them an almost impossible burden of proof. A bad actor can easily claim he or she was just experimenting or making a parody, thereby leaving the law toothless. Meanwhile legitimate artists face constant legal uncertainty. It fails to catch the malicious operator yet successfully intimidates law-abiding citizens.
Beyond these specific flaws, we must look at the wider picture. Now is simply not the time to be pushing yet another piece of legislation into the mix. A number of Bills that deal with this issue are already before the Dáil. I think a Bill from Deputy Malcolm Byrne passed Second Stage in the Dáil in March 2025. In addition, massive legislative frameworks are under way at the EU level. Instead of rushing this Bill through, we need to stand back and wait to see how EU laws come into effect so we can respond appropriately. This is especially true considering that the Government has yet to actually transpose the relevant EU legislation into our Irish law. Piling another unco-ordinated domestic Bill on top of an unfinished European framework will only lead to legal chaos and duplicate rules that contradict one another.
The structural flaws multiply when we move from criminal law to civil remedies. Section 8A lets the Circuit Court issue identification orders to force Internet companies to unmask anonymous uploaders so victims can sue them. It sounds like common sense on paper but it completely ignores the harsh reality of our Judiciary. Our local courts are already facing severe backlogs. Flooding the Circuit Court system with highly complex emergency digital recovery applications, where local judges must balance intricate European data privacy rights against a plaintiff's claims, will cause a total administrative bottleneck. Without dedicated specialised judicial resources, this reactive policy tweak will break an already strained system. Even if a victim secures an order, we run directly into a wall of international borders.
Section 2B attempts to criminalise inducing someone to create a deepfake, and explicitly states that the person being induced does not need to be in Ireland. This is an illusion of security. Irish courts have strictly limited jurisdiction over non-citizens operating abroad. Successfully identifying, extraditing and prosecuting an anonymous user hosted on a foreign server or residing outside our borders presents immense international law hurdles that this text simply glides over. It offers empty comfort to victims but does little to solve the cross-border reality of digital abuse.
We must look at the most radical tool in the Bill. Section 6B empowers Coimisiún na Meán to seek High Court orders to completely block access to an entire online service or application store within Ireland. This is basically using a sledgehammer to crack a nut and places this Bill in direct conflict with European Union law. Under the EU Digital Services Act, online platforms operate with a unified Single Market compliance framework with specific liability exemptions. A single member state acting unilaterally to institute a nationwide Internet block risks violating EU law and disrupting the Single Market. More fundamentally, blocking an entire platform because a subset of users abuses it constitutes a form of mass censorship that may directly challenge Article 40 of our Constitution and Article 10 of the European Convention on Human Rights. It punishes thousands of lawful users, stifles digital commerce and sets a dangerous precedent for State-sanctioned Internet blocking.
True protection for victims does not come with sweeping Bills that conflict with EU law, overburden our courts and threaten free speech. It comes from precise legal drafting, explicit protections for legitimate expression and robust enforcement mechanisms that work across borders. As it stands, the Bill is slightly flawed. I commend it, I will be supporting it and I will be putting down amendments on the next Stage but I welcome the sentiment of my colleagues' Bill.
Comment on this
I also commend Senators Flynn, Ruane, Higgins and Black on bringing this important legislation forward. I am delighted to hear that the Government is not objecting to it, which is great, because it has been brought up across the floor whether it is on the Order of Business or as a Commencement matter. It is an issue that deserves our attention because, as has been said before, technology sometimes does move more quickly than our laws can move. Technology moves so fast that sometimes it is impossible for the law to keep up with it. I know it would be nice to have it there before something happens but that is not always possible so I have to commend my colleagues on getting this in so quickly.
AI has brought enormous benefits whether it is in work or when one rings up a company. Going through the AI system can sometimes be quicker than speaking to a person. However, AI brings a lot of the other side where we do not fully understand what can be done, so there are two sides to AI. In no way am I standing here bashing AI. There is a place in society for it. We just need to have the mechanisms to protect people.
One of the most alarming developments is the creation of disturbing images. For a person who is targeted, the consequences are absolutely devastating - anything from humiliation to distress, anxiety and depression to reputational damage to the profound invasion of his or her privacy. Unlike many other forms of abuse, once an image goes online, it is almost impossible to rein it back in or gain control over it. What makes this particularly concerning is that a person no longer needs to have taken that intimate image, photograph or video for someone to exploit him or her. The photograph can now be taken from a social media account and potentially be manipulated and used for something completely degrading entirely without his or her consent.
This is an issue that disproportionately threatens women and girls but it can happen to anyone. It can happen to boys and men. It is quite scary how easily it is done. Consent must remain at the heart of this conversation. A person's image, identity and dignity are not public property. Having a photograph online does not give everybody or anybody the permission to manipulate it into something sexually explicit. I particularly welcome the provision in the Bill that creates specific criminal offences for the creation of deepfake intimate images without consent and where there is an intention to cause harm.
Importantly, the Bill also addresses those who induce or attempt to induce another person to create these images. That is significant because responsibility must not solely lie with the person who physically generates the image. Those who commission, encourage or facilitate the abuse must also be held accountable.
It is fantastic to have the provision in the Bill that allows victims to seek identification orders through the Circuit Court. This is very important because 90% of the time, stuff is anonymous online and it is impossible to figure out if one is being targeted and who is targeting one. That is a very important provision to have in this Bill.
Another particularly important aspect is the provision allowing the courts to order access to services to be blocked where those services are being used to create deepfake images. Online platforms cannot simply wash their hands of their responsibility.
The responsibilities of the online platforms have been discussed at Oireachtas committees and in the Dáil. They have a huge role to play in preventing this abuse and protecting their users. Of course, any measures taken must be proportionate and must respect the rights of users. I welcome the safeguards and judicial oversight being provided in this Bill.
Ultimately, this legislation is about more than AI. It is about dignity, consent, privacy and the right of every individual to live free from harm and exploitation. We have seen how quickly technology can move ahead of legislation. We cannot afford to wait until somebody else suffers serious harm before we act which is why I am delighted to see this Bill before the House today. Our laws must recognise that abuse is not just physical. Our laws must protect people online as well, given the damage that online abuse can do. Again, I commend Senators for bringing this forward. I am more than happy to support it and am delighted to hear that the Government is supporting it too. It is very important legislation that will enable us to respond quickly to the way the world is moving. It is so important that we keep pace.
Comment on this
Ar dtús ba mhaith liom mo bhuíochas a ghabháil leis na Seanadóirí Neamhspleácha as an mBille seo. Is Bille tábhachtach é agus mar sin beidh mé ag tabhairt tacaíochta don Bhille inniu, ach tá cúpla ceist faoin mBille freisin agus beidh mé ag caint faoi na ceisteanna i rith m’óráide. I thank Senators Flynn, Ruane, Higgins, and Black for introducing this Bill.
As has been indicated, the Bill proposes to amend Coco’s Law to criminalise Al-generated non-consensual intimate images and videos. I agree with the objectives of this Bill, which is why I secured Government approval yesterday not to oppose it today. Non-consensual image nudification can be deeply harmful to a victim even if no real nude photograph ever existed and even if such image is never shared with anybody else. The reality is that such actions represent a violation of bodily autonomy as the victim loses control over how their image and identity are presented to others. They reduce persons - generally women - to sexualised objects without their permission. Even if such an altered image is never shared, knowing that others have it in their power can cause significant distress to a victim. There is a common misconception that if it is not real then nobody is harmed but in truth the harm comes from the humiliation, fear and loss of control experienced by the victim. In addition, evidence suggests the generation of non-consensual intimate images disproportionately targets women and children. I say this not to diminish the fact that men can also experience this behaviour but to highlight the reality that it is mostly women and children who are targeted by these actions. They reinforce existing patterns of sexual harassment and gender-based abuse and have the potential to discourage female participation in public, professional or online spaces. The Bill therefore aligns with the Government’s commitment to a zero-tolerance approach to domestic, sexual and gender- based violence which is a priority for me and for Government. Therefore, we will not be opposing this Bill today.
However, as I said in my introduction, I have a number of questions with the Bill as it currently stands. The Office of the Attorney General has advised that the definition of “intimate image” in the 2020 Act is sufficiently broad to capture images created or altered by Al tools. An amendment in this regard is therefore not necessary. With respect to the new offences, it is not clear how the creation of a deepfake image without publication or threat of publication could cause harm to another person unless that person was made aware that the image had been created. Furthermore, the offence of inducing a person to create a deepfake image essentially makes it an offence for a person to tell another person that a third party has consented to a deepfake image of that third person being created, where such consent has not actually been provided and where there is intention or recklessness in respect of harm caused. It is not necessary for any deepfake intimate image to have been created nor for the inducement to have been directed at any specific person, which appears problematic. In this context, officials in the Department of Enterprise, Tourism and Employment have noted that the EU Artificial Intelligence Act expressly defines and regulates deepfakes, in terms of Al system transparency obligations. Alignment with definitions established under European law would be preferable to ensure consistency of application. Multiple definitions in national or EU law for similar terms such as deepfake are not considered helpful. In addition, officials in the Department of Culture, Communications and Sport have advised that the proposed access blocking order in section 7 of the Bill would allow Comisiúin na Meán to seek an order before an investigation has actually concluded. This would appear to run contrary to the rules of fair procedure.
Having said all of the above, there are a number of developments that have taken place, particularly at EU level, which are important to share with this House, and which may affect the approach to be taken by Government in future legislative stages concerning this Bill. I should also add this is not the first Private Member's Bill introduced in the Houses that aims to criminalise the non-consensual nudification of images. In fact, it is the third such Bill.
I will first turn to the current legal framework. As Senators know, through Coco’s Law lreland established a comprehensive regime to address intimate image abuse, including offences for distributing, publishing or threatening to distribute intimate images without consent. The most serious offence is punishable by up to seven years’ imprisonment. Coco’s Law complements the Child Trafficking and Pornography Act 1998, which criminalises the production, distribution, and possession of child sexual abuse material. The other legislation that is very relevant to this debate is the directive on combating violence against women, the first EU-wide law specifically addressing violence against women and domestic violence. It was adopted in May 2024 and transposition is currently ongoing in my Department. One of the directive’s most significant achievements is the creation of EU-wide minimum rules on several forms of cyberviolence, including the non-consensual production and sharing of intimate images. I anticipate the policy objective underpinning this Bill will be achieved through the transposition of this directive.
It should also be noted that since I addressed the Dáil earlier in the year in the context of the first Private Member's Bill introduced to address this issue, important developments have taken place in relation to the EU Artificial Intelligence Act. The EU digital omnibus on Al, which entered into force in July 2026, is a set of amendments to the Al Act with a primary purpose to simplify the implementation of the Act. However, the adopted text also bans Al systems that generate child sexual abuse material or create images, videos or audios depicting an identifiable person’s intimate parts, or sexually explicit activities, without their consent. Providers will not be allowed to place these systems on the EU market unless they come with adequate technical safeguards to prevent the creation of such material. This outright prohibition also applies to deployers using them for this purpose. At a domestic level, the Al Act 2026 gives legal effect to national implementation measures to establish a regulatory framework to meet national obligations under the Al Act. This includes the designation and empowerment of national competent authorities to regulate prohibited Al practices such as the one we are discussing today.
I say this because Ireland has been a strong and determined advocate at EU level to prohibit Al systems which permit this harmful behaviour. As technology evolves, so too, inevitably and unfortunately, do the methods used to target, exploit and abuse women and children. Our response must evolve just as rapidly, ensuring that our laws reflect the realities of abuse in the digital age and provide effective protections for victims. In line with the programme for Government and throughout Ireland’s Presidency of the Council of the EU, Ireland is actively engaged in advancing European measures aimed at combating child sexual abuse and exploitation, including efforts to ensure that evolving forms of online abuse are effectively addressed in legislation. I was pleased to welcome the recent agreement on a new directive updating the EU’s criminal law rules on child sexual abuse and sexual exploitation of children which was a key deliverable of Ireland’s Presidency. This agreement was another big step towards helping us to better protect and support our children both online and offline, and bringing to justice perpetrators, particularly those who think they might evade justice because they act behind a screen or use new technologies to target children.
While many of the new measures are already addressed in domestic legislation, the criminal law (sexual offences, domestic violence and international instruments) Bill 2025 will modernise our legislative framework. Crucially, this will include replacing the term "child pornography" with "child sexual abuse material" in legislation, further recognising that such material is evidence of abuse and exploitation.
We are also supporting research to better understand and combat emerging online harms, including AI-generated imagery, deepfakes and other forms of technology-facilitated abuse.
Many of these priorities are also reflected in our Presidency engagement, where co-operation is key to strengthening our collective responses to violence against women and children. Preparations are continuing for the transposition of the directive I mentioned earlier. This landmark legislation will strengthen protection for victims, including in relation to technology-facilitated forms of abuse, and will support a more consistent approach to preventing and combating violence against women and domestic violence across the EU.
Over recent months, we have advanced a number of legislative measures to strengthen protections for victims and survivors and to support a more effective justice system. In July, Valerie's Law and Jennie's Law were enacted, thereby enhancing protections for children affected by serious family violence and increasing transparency in relation to domestic violence convictions. Reforms to the disclosure of counselling notes and other sensitive records will better protect victims' privacy while safeguarding the interests of justice. The introduction of new offences targeting the exploitative practice of "sex for rent" further demonstrates our commitment to protecting vulnerable people from coercion and abuse.
In addition, work is continuing on legislation to strengthen consent laws in sexual offence cases, including how belief in consent is assessed. In July, I secured Government approval for legislation to criminalise the production, distribution and possession of extreme and violent pornography, reflecting growing concern about its harmful impact, particularly on young people, including the normalisation of sexual violence and the reinforcement of harmful attitudes towards women and relationships. Taken together, these reforms represent significant progress in strengthening protections for victims and survivors.
I commend the Independent Senators and all the speakers in the House today on their determination, which I and the Government share, to eradicate the blight of domestic, sexual and gender-based violence. This is a challenge that requires sustained commitment across Government and society and one to which we remain fully committed.
Comment on this
Cuirim fáilte roimh an Aire. I too commend Senator Flynn and her team, including Corey, and all the other Senators in the Civil Engagement Group on bringing this forward. We in the Labour Party are delighted to support it.
As has been said, behind every manipulated image is a real person. The fact that anyone can take an ordinary photograph of a person and turn it into sexual abuse material without that person's knowledge or consent beggars belief. Just because the image is fake does not mean that the fear, humiliation and damage to someone's life is not very real. Sadly, all of us have heard the deeply upsetting stories from people who have experienced this kind of abuse. We have heard about the violation of dignity, the fear and the terror of humiliation that comes from these cases. No one should have to live with that threat. No one should have to live in fear that their image or identity can be abused, manipulated or weaponised without consent. No woman should have to face the threat of having her image, reputation or dignity violated in this way. I am delighted that the Government is letting this Bill go through because it is our job as legislators to respond to new threats to the safety and dignity of our fellow constituents. That is why this Bill is so important.
The Labour Party, specifically my colleague Deputy Alan Kelly, published our own Harassment, Harmful Communications and Related Offences (Amendment) (No. 2) Bill nine months ago. That was directly after the Grok scandal, whereby a freely available AI tool on one of the largest social media platforms was used en masse to target women and girls, churning out many thousands of abusive deepfake images of real people over a period of weeks. Across the Opposition parties - I know Sinn Féin has brought in a similar Private Members' Bill - we are bringing forward constructive solutions to the problem of deepfake image abuse.
It is great that this Bill is passing to the next Stage, but we really need to see the Government act on this now. The online world and AI images are not getting any better; they are getting worse. The serious proposals brought forward on this ensure that we will do right by women and girls and anyone who may be a victim of deepfake abuse. I commend the provisions on anonymity, bail and domestic violence aggravation, which are really important aspects of this Bill. I want the Government to ensure that it will not stall on it, that it will listen and react accordingly and not always be one step behind and that this will progress quickly. As has been said in all the interventions here, things are getting a whole lot worse and are not getting any better. I wholeheartedly support this and I thank Eileen.
Comment on this
I commend my colleague Senator Eileen Flynn and Corey, her policy adviser in her office, on their dedicated work over the past 18 months alongside the Office of Parliamentary Legal Advisers in bringing this very important legislation before this House. Senator Flynn has set out the clear structure of the Bill. It fits together to offer some protections for victims of image-based harassment, mainly women, girls and young people. These victims cannot afford to wait while technology moves so rapidly ahead. Sadly, we have already been too slow to address these technological advancements. AI tools that remove clothing from photographs are no longer rare or difficult to find. Domestic violence organisations, including the rape crisis centres, our youth workers and teachers tell us that such tools are commonly being used across the country in secondary schools, universities and workplaces and even to control women trapped in abusive relationships. It is horrific to think of it. Some of us have young children and young grandchildren. I have two beautiful grandchildren, and it is awful to think what is ahead in this area.
Right now, when a woman discovers that an ordinary photograph of her has been taken from social media and used to manufacture a fake intimate image, our legal system fails her at the very first hurdle. If that fake image has not been shared publicly yet, the person targeted is left in limbo with no recourse to legal remedies against the perpetrator, the platform that hosted the image or the AI tool that created it. We cannot ask victims and An Garda Síochána to fight this battle one case at a time while rogue websites and smartphone apps continue to profit from digital abuse. By classifying this material as harmful online content under Schedule 3 to the Broadcasting Act 2009 and giving Coimisiún na Meán the power to apply to the High Court for access-blocking orders, this Bill creates a mechanism to shut these reckless operators down. Where an online service knows, or reasonably should know, that it is being used to create illegal deepfake intimate images and fails to take proper steps to stop it, the High Court can order internet service providers and smartphone app stores to block access and software downloads right across the State, giving specific legal priority to protecting children from harm.
Having a realistic fake intimate image created of you without consent is a terrifying prospect for many and creates a deeply troubling power dynamic, particularly for women and vulnerable young people. It can be used to blackmail, to silence or to control in a coercive relationship. In some cases a person does not even generate the image on their own device; instead, they upload a photograph to an anonymous online message board and ask, prompt or pay a stranger online to remove clothing from that image. These novel technologies are being developed and implemented so rapidly that victims of their use are left with absolutely no protection. The Bill recognises the seriousness of the creation of non-consensual deepfake intimate images by making such activity an offence.
Just as importantly, the Bill recognises how deeply connected this abuse is to domestic violence and coercive control. By updating section 40 of the Domestic Violence Act 2018, this Bill requires judges to treat an intimate relationship as a mandatory aggravating factor at sentencing. We must empower victims, mainly women, girls and young people, to come forward when they are subject to this type of abuse and we need to do our utmost to protect the victims and put proper provisions in place to protect and support them. I am delighted to hear that the Minister and colleagues across the House will support this Bill today.
Comment on this
I offer a special thank-you to Senator Flynn. She has been really consistent on addressing issues around domestic, sexual and gender-based violence. She has done it with the child sexual abuse material legislation and this legislation, and it is really important we have such consistent voices in this House.
I am delighted to hear the Government is accepting the Bill. It is a loophole we currently have in our legislation. The growing use of artificial intelligence, AI, to create these deepfake intimate images and videos represents such a disturbing development in the abuse of technology, and we have to deal with it.
The images and videos we are talking about are not harmless or online jokes. They are forms of sexual abuse and are designed to humiliate, intimidate and violate another person's dignity. We only need to look at the experiences of two women who have publicly spoken about this to see that.
Cara Hunter, who is an SDLP MLA in the North, has been very courageous and consistent in speaking out about the deepfake sexual video that was created of her and circulated during her election campaign. It is the stuff of nightmares for many people seeking election or for elected representatives that this could happen to any one of us. The consequences for her went beyond the online world. She has spoken about men approaching her in the street following the circulation of that material or laughing at her when she was canvassing. I cannot imagine the fear or sense of humiliation and the impact this would have had on her as a young woman in the public sphere and on her wider family when she was just trying to do her job and represent her constituents.
Her experience also raises serious questions about the additional barriers women face when they are trying to enter and remain in public life. I also reference Gráinne Seoige, whom we have all seen speaking out about this issue and the deepfake intimate images circulated during her election campaign. She described the trauma of discovering what had happened to her and how, for months afterwards, she wore a surgical mask when shopping because she was afraid of being recognised. Not even wanting to leave the house with your face visible is a severe level of trauma. She felt she had to disguise herself when simply trying to go about her everyday life because of a technology that was used to violate her dignity.
Neither of these women got the support they needed when they needed it and the images were not taken down in a proper manner. Despite doing everything they could, they have never found out who made those images and no one has been charged. They are women in positions of power with access to decision-makers. These are women with a degree of privilege compared with many others. If a woman who has a platform and a public voice cannot even get the help she needs, how difficult would it be for an everyday woman to get access to justice? We have much more vulnerable women of all different ages that this will be happening to and their recourse is very limited.
However, their experiences and courage in talking about it publicly have shown us the enormous amount of harm that occurs when an image is created or shared. It follows the victim into their homes, their workplaces, their relationships and everyday lives. It is a form of the most grotesque online bullying. When we think about it, this is not just online; it is in every reality they live in. Senator Flynn and the rest of the Civil Engagement Group have brought forward a very important amendment that is crucial in protecting people from online violence. It recognises that consent, dignity and privacy must be protected in the digital world, just as we hope to protect them in the real world.
Technology is used to abuse, intimidate and silence and we have to stop that across all forms of technological abuse. Today, I have used examples of public figures but I reinforce this is something that is insidious. It is happening in every community to people of all ages, including those under the age of 18. It is gendered and specifically used most egregiously against women and girls. If we do not look at the gendered impact of that, then we are missing a key aspect.
Again, I commend Senator Flynn, who has worked very hard on this Bill. She has been a tireless advocate when it comes to domestic violence and sexual and gender-based violence. I am very proud to be sitting with her today to provide that support.
Comment on this
First, I thank the Minister, Deputy O'Callaghan, and his Department for not stopping this Bill from moving forward today.
As not just a woman but as a woman from the Traveller community, not for a minute did I think when I came in here nearly seven years ago that I would be part of two major changes in legislation in this country with one to protect our children when it comes to child sexual abuse images and materials online. I will not always be here to protect my two wee girls who are now five and seven years of age. They are going to grow up, and I want the multinationals such as Facebook, Meta and TikTok to protect them.
I am privileged to sit on the gender-based violence committee, along with my colleague here beside me. We are the only two members from the Opposition from the Seanad who will be on the committee, and that is brilliant. However, as the Opposition, we always have something to give out about but it is regretful that the other Senators who sit on the gender-based violence committee are not here to share a few words on how AI is having a major impact on women's lives. Thanks be to God, I have not been impacted by it personally but by God, I have got many horrible comments about my image on social media in recent years.
However, in the past 18 months, after the Minister for justice, Deputy O'Callaghan, accepted our child trafficking and sexual abuse material Bill, we went back to the office and thought about what we could do now to go to the next part. This was before the scandal came out about Grok and AI was only really coming to the surface around women and children's images. I acknowledge the Minister and other Senators in the Opposition have spoken about the EU laws but those laws do not go far enough. For us, it is a simple piece of legislation.
Again, I stress we have worked on this legislation with the OPLA, which has put in a lot of work over the past 18 months. It is a service we know the Department of justice trusts. Many Senators are right that there are a few changes that need to happen in the Bill to make it best practice, because if you are going to take on a job or change something, it is better to do it right the first time round. We welcome the changes the Minister and Department are looking to make with the Bill, because it is important it works effectively.
I thank the Irish rape crisis centres, Children's Rights Alliance, Safe Ireland, Pavee Point and the other organisations that did not come in behind us but instead came in beside us today to talk about this very important legislation to be passed in this country. I am extremely proud and I hope to keep going forward in bringing in legislation that is important for women and children in this country. As I said, I have two wee girls of five and seven who I will not always be here to protect. There are many mothers who worry about their kids going forward. We have to look at it and say this is the job of the State and of the multinational tech companies. With this legislation, we are making them take responsibility. We are giving a platform for victims and a pathway to justice for any person, even if they are never a political representative; it is also for ordinary people on the ground. It is a pathway and is not just the bare minimum. It is 18 months of hard work, from Corey, the OPLA and myself.
I thank the Minister and the Department of justice for accepting our legislation here today.