We have apologies from Deputy Mythen. Deputy Ó Cearúil is chairing the Gaeltacht and Gaeilge committee and is hoping to join us but he sends us his apologies for the start of this meeting. There are obviously the usual constitutional limitations but I am quite certain the witnesses are very familiar with those. I welcome the Minister of State, Deputy Smyth. Her presence was requested at short notice but she was keen to come and join us today and I appreciate her being able to adjust her schedule for us.

We are covering the topic of AI and public sector programmes but the committee felt it would appropriate, given some of the recent issues concerning AI, to invite the Minister of State to update the committee, especially in light of recent meetings and the decision of the European Commission.

I invite her to deliver her opening statement.

I thank the Cathaoirleach and colleagues for coming in on this wintry day to be with us and discuss this very important topic. I am joined by my esteemed colleagues from the Department of Enterprise, Tourism and Employment Ms Jean Carberry, assistant secretary of the digital division, and Mr. Declan McCormack, principal officer for AI and digital regulation.

As Minister of State with responsibility for artificial intelligence, I welcome this opportunity to appear again before this very hard-working committee. The online safety matters that have so acutely come to the fore in recent weeks have highlighted the importance of having an effective legal and regulatory framework to deal with issues such as deepfakes and consent. They also underscore the importance of the work of the Government, my Department and this committee in ensuring the necessary guardrails are in place to build public trust in AI. I share the deep concerns that have been raised by members of the committee and I thank the members who took the time to come to the debate we had in the Dáil on that. I understand we will also have a debate in Seanad Éireann and I look forward to engaging with Senators then too. I share the deep concerns that have been raised by members, and the public in general, about reports of harmful and illegal content being generated and disseminated through the use of Grok on the X platform.

This is a very serious issue that affects not just Ireland but all EU citizens. This Government and the European Commission have made it clear the use of AI to generate offensive material or to abuse or harass any individual is absolutely unacceptable. The recording, distribution or publication of intimate images without consent is illegal and the generation of child sexual abuse material is illegal. An Garda Síochána has stated its full commitment to supporting victims and has numerous investigations under way to determine if there have been breaches of the law, including under the Harassment, Harmful Communications and Related Offences Act and the Child Trafficking and Pornography Act.

If a member of the public is concerned about images shared online, there are three steps to take. The first is to report it to hotline.ie to get the content removed. The second is to report it to An Garda Síochána to hold those responsible to account. The third step is to report it to the platform, and if it fails to act, to notify Coimisiún na Meán. The Digital Services Act, DSA, is a crucial piece of Ireland’s online safety framework. It places an obligation on online platforms to act on reports of illegal content. Under the DSA, the European Commission is responsible for the oversight of very large online platforms’ compliance with their requirements to assess and mitigate risks that their services may create in relation to the proliferation of illegal content online and the protection of fundamental rights, including the protection of minors.

Coimisiún na Meán, in its role as Ireland’s online safety and media regulator and digital services co-ordinator, is actively engaging with the European Commission in this context, in relation to the use of Grok on the X platform. Coimisiún na Meán has highlighted to the Commission the urgent nature of the issues and the importance of an effective and timely regulatory response. I welcome the European Commission’s decision yesterday to open a formal investigation into X, following Coimisiún na Meán’s intensive engagement with it on this issue in recent weeks.

European and Irish law puts clear responsibilities on online platforms relating to illegal content and legislation is underpinned by a pan-European system of regulatory enforcement overseen by the European Commission and national regulators. Through my engagement with Coimisiún na Meán in recent weeks, I have every confidence it is ready to play its part in the investigation. Building on the work of the regulators, an urgent cross-government response has been mobilised in relation to the issues, including the relevant Ministers and the Attorney General as well as Coimisiún na Meán, to ensure a full understanding of the legislative and regulatory framework around this issue. Alongside this, and as recommended by the AI Advisory Council, which I thank for acting so swiftly on this and providing us with a detailed document and four clear recommendations, the Government will explore the potential to expand the list of prohibited AI practices under the AI Act and explore options to strengthen victim reporting and awareness raising. In that regard, I have asked my officials to liaise with the Commission on the potential for the AI Act to be amended to provide protections against the generation of images and deepfakes without consent and the generation of content constituting child sexual abuse material through AI systems.

The opportunity AI offers is immense, but so is our responsibility to get the balance right between innovation and protections to ensure AI is trustworthy, human-centric and used for our greater good, for citizens and for business, in the economy and in wider society. That is why we have been actively working to put the enablers and guardrails in place by increasing awareness and understanding through trusted independent voices, such as the AI advisory council, implementing the EU AI Act, establishing a new AI office of Ireland and making Ireland’s voice heard in international AI governance forums such as the EU AI board. The recent misuse of AI to generate appalling deepfake images is a prime example of how trust can be eroded and why governance is so important. Our work is vital in cementing Ireland’s position as a centre of regulatory excellence and as a strong voice in Europe advocating for well-designed digital regulation that is fit for purpose to protect individuals’ safety and fundamental rights while promoting the adoption of trustworthy AI and enabling innovation.

I look forward to a constructive engagement and, of course, members' suggestions and observations. Go raibh maith míle agaibh.

Go raibh maith agat a Aire Stáit. As we have quite a number of speakers and the Minister of State can only stay with us until 1 p.m., I ask people to take four minutes for questions and answers. I am aware that is tight but a lot of colleagues are coming in. Please try to be as succinct as possible. Deputy O'Rourke is first.

I thank the Minister of State. Is it her analysis that there was a failure of regulation here or that the regulatory powers are insufficient? It strikes me that there is a failure of regulation and that in essence the Digital Services Act provides tools that were not acted on in advance. In fairness to the Minister of State, she has done it with a bit more compassion and understanding, but what we have had in the last couple of weeks has been a complete failure of regulation. Is it not the case that we should not have been in this position at all? This type of facility should not have been sanctioned and operationalised in the first instance, in Ireland or in a number of other jurisdictions. Is it the Minister of State's analysis that it is a failure of regulation?

If it is that, what went wrong?

I thank the Deputy for his questions. We have had this controversy roll on for a number of week and for the Deputy, me and anybody watching or any of our citizens, a day of this continuing, or the availability or capability of the app, is a day too long. In an ideal world I would like to be able to shut it down immediately when we see something like that happen.

How did it go live? How does something as potentially harmful as this even go live without oversight?

As the Deputy will know, Grok is an X AI tool and Grok has, to my understanding, been around for quite some time. When X decided to use the X platform to raise the profile of it, which we also have done by having a public discussion around it, that provided the capability to people, bad actors we might say, to use it in a way that is inappropriate and is harmful. However, we have a regulator, Coimisiún na Meán, we have the Digital Services Act and the regulator within the Commission in Europe, and our regulator, working with An Garda Síochána based on the reports and the complaints received, has reacted as quickly as it can to work with the Commission to have this shut down.

I engaged with X immediately. I met with them. Saying they apologised might be a step too far, but they did concede the fact that when they designed this app it was not intended to be used in this particular way. My argument to that would be that when somebody designs a tool or an app that has this type of capability the safeguards and due diligence should also be matched with the capability of the technology in the first place.

That is my point. Is there not provision for an independent risk assessment for such tools before they go live? I know there is an annual report. The risk in all of this is that we do not know what the next example of this will be. We cannot leave it to people like Elon Musk or companies more generally to self-regulate.

The DSA does require platforms to carry out that risk assessment and due diligence on them. The Deputy and I can sit here and say obviously and evidently that was not done. I cannot pre-empt the outcome of the investigation of the Commission, but I assume that will be what the investigation turns out.

Given the time constraint, I am going to fire what I think are the most important questions from the group I was going to ask because I will not get to them all.

We are looking at the difference between immediate enforcement versus a future-facing response that I feel is probably unfolding rather than the immediate enforcement. Regarding Coimisiún na Meán, we know it has the power to investigate and sanction under Part 8 of the Online Safety and Media Regulation Act. Has Coimisiún na Meán opened a formal investigation into X and Grok for breaches of the online safety code, and if not, why not?

Regarding the Data Protection Commission, we know that under section 134 of the Data Protection Act 2018 the DPC can urgently restrict or suspend data processing to protect fundamental rights. Why has the DPC not exercised these powers in the immediate in relation to X or Grok despite clear risks under GDPR Article 35?

Regarding consumer protection, has the Competition and Consumer Protection Commission assessed whether X or Grok terms unfairly shift liability onto users for harms caused by the platforms own AI systems, and if not, why not?

I will skip a few of these questions because of time and move on to the final question.

I encourage the Senator to give me all the questions, because I will take note of them and try to come back on each of them.

In relation to corporate criminal liability and Coco's Law, does the Government accept that where AI creation and publication occur simultaneously, platforms like X care bear direct criminal liability and, if so, what action is being taken?

What concrete deterrent can be established in Ireland because we have people in Ireland using the platform in this way? I am thinking in terms of deterrents which is obviously more of a preventative, future-facing piece.

I thank the Senator. She has given me very broad and insightful questions. Some are my remit, and some are not, but I will come back to her.

In response to the first question and Coimisiún na Meán, my understanding is it has an investigation underway which is collaborating with the Commission and will be feeding the information to the Commission. It feeds in and informs the investigative steps that will be taken by the Commission as well. Coimisiún na Meán is named as a party formally associated with the Commission's investigation and will be playing an important role. It will be gathering the evidence and sharing that with the Commission for its investigation.

I encourage the Cathaoirleach to bring in the Data Protection Commission in the future to get a full, clear picture to see what its role is. I cannot answer the question relating to the DPC directly for the Senator, but I would be keen to meet with the DPC in the very near future in case something like this happens again.

With regards to the Consumer Protection Commission, the Senator is quite right. It has been my sense from dealing with X that there is a sort of shift in terms of bad actors and user manipulation which I do not buy if the technology is designed to have this type of capability. It should not be designed in that instance in the first place. If there is a sense it can be manipulated in anyway the safeguards should be in place to not allow that to happen. The law is the law in Ireland. There are laws in the European Union. That is why the DSA was created. Having met with the Attorney General when this first broke, I am confident in the three laws we have in place.

To the question about the AI generated images, having gone line-by-line through all the provisions we have in place with the Attorney General, my understanding is that AI generated images are prohibited when we are talking about children and this use of this type of technology.

Anything I have not answered for Senator Ruane my officials have taken notes on and we will try to come back with specific answers. I encourage the committee to have a hearing with the DPC as it could be useful to us all.

I thank the Minister for her comprehensive opening statement and the information she provided. I commend her on the strong line she took on this immediately. She was very clear on the airways articulating to the public that this was not acceptable and was, in her view, prohibited by both European and national law. The Minister has gone through with us what the next steps are to ensure, if a breach of EU regulation has occurred, there will be sanctions from the EU Commission and that those responsible will be held accountable.

To summarise, An Garda Síochána is investigating whether any criminal actions have taken place on an individual basis with these images that were generated. The EU Commission is examining if regulations have been breached under the DSA and, on foot of the AI Advisory Council's recommendations the Minister is reviewing legislation to see if there are any gaps and if there is anything else we need to do.

Regarding the Minister's interactions with X and its representatives, I want to ask about the manner in which they responded to her, the speed with which they came back to her and the level of engagement she got there. Was the Minister satisfied with the responses she was getting from X as the controversy unfolded?

I thank the Senator for her questions. I met with the representatives from X quite quickly. It was an online meeting. We had their EU policy representative there, we had some of their legal team and some of their regulatory team as well. As I said, I found their perspective skewed. The language they were using about user manipulation and bad actors does not sit well with me. I understand what that means. However, this is a huge company with huge financial resources behind it that has the design team and capability to design this kind of technology in the first place. We have very clear laws in Ireland, and we have laws under the DSA too, that guardrails and due diligence should be in place with any technology that is going to affect European citizens.

I heard executive vice president Ribera this morning with David McCullagh. I look forward to meeting her this afternoon in the sense that we are all on the one page within the European Commission. We have values within the European Commission that are about the protection of human dignity. This is the most dehumanising tool that could possibly be on a plain and easily accessible platform that we have seen so far. It is incumbent, having gone through it all with the Attorney General that we have the laws in place, that enforcement is key. An Garda Síochána has been before our committee to talk about the number of cases it is taking. My focus and emphasis are on X and Grok itself, not so much the users. Of course the Garda will take care of the users, however, I feel the most critical element for us as legislators is to ensure the enforcement piece at a European level to ensure this does not happen again.

I fully agree with the Minister of State; that is where her attention should be focused and rightly is so. I have concerns from a response perspective. I did not think X responded as quickly as it could have to a Government Minister bringing this to its attention. Is there anything we can do in the future to ensure a speedier response? Are there handbrakes we can pull in a situation like this where there are question marks over the legality or potential misuse of an application?

I met Coimisiún na Meán on a couple of occasions in relation to that because we want to know domestically that we can pause the capability of this technology being proliferated among our citizens. We also want to know the European Commission can deactivate that kind of capability. I am confident it can when necessary.

I welcome the Minister of State. It is a very timely intervention. From research, we know this technology is generating approximately 6,700 suggestive images per hour currently. About 34 million images per day are created by AI chatbots. Unfortunately, the horse has very much bolted. The Minister of State in her opening statement outlined the steps people who come into contact with this and who it happens to have to undertake. They can report it on hotline.ie, to An Garda Síochána and to the platform. If it does nothing, it can be raised with Coimisiún na Meán. We all know the platforms are not doing enough. That is where the regulation and enforcement have to come in. We welcome the announcement yesterday. Lots of stuff is going on at the moment in relation to AI. We are looking at Grok right now, but is there a mechanism in place to tackle future applications in terms of the criteria governments will use to determine whether additional AI applications should be restricted or prohibited moving forward under the EU Act.

I thank the Artificial Intelligence Advisory Council, an independent group of experts in Ireland headed by Dr. Patricia Scanlon. It met very quickly to look at the legislation we have in place around AI. It came up with four clear recommendations on what we can do to prevent something like this happening again. One was to expand the AI Act's list of prohibited practices. We already have that in train to look at how we can broaden our prohibited practices. I hope throughout our Presidency we can have that as one of the key pieces. Another was to support victim reporting and to look at guidance on how we can protect the public and the piece on key messaging to create public awareness. It also suggested an information campaign to explain to the public the use of AI, the creation of non-consensual intimate images and where automatically shared, the potential criminal offences. There is work to inform the public particularly within schools and young people. That generation is tech savvy and it can cause harm. Coimisiún na Meán has done some work on this as has An Garda Síochána. One can see their ads on mainstream media and RTÉ about how the sharing of images or even the attempt to share images can be a criminal offence. We have more work to do on that. Another is to create a national taxonomy of online harms that includes AI-enabled harms to standardise reporting and policy. The key piece in terms of legislation is to expand the list of prohibited practices. I hope that will ensure this will not happen again.

We need to look at preventing the use of AI for political manipulation moving forward because of the focus on a particular area, the recent public outcry and a lot of public representatives leaving certain platforms. That has to be an important factor in anything that comes out of this.

I will take those suggestions on board. I think everybody in this room feels very strongly about democracy because democracy is what got all of us here and having a variety and diversity of views where we can reach consensus. There was nobody in the Upper or Lower House who was not completely appalled by what happened. It gives us room and space to ensure we drill down into the detail, as I did with the Attorney General, to make sure the protections are in place. Enforcement is key but equally the main message today is about information and ensuring public information and information in schools is strengthened.

I thank the Minister of State for her engagement with this committee particularly at this moment in time. The Minister of State met X. Did she meet its representatives in person?

It was an online call.

How many people from X were present in the meeting?

Three.

Where they people work and are based in Ireland?

One in Brussels and two in London, if I am correct.

Does the Minister of State know how many people X currently has working in Ireland?

I do not have the exact figure but I think it is between 120 and 150.

In the course of the conversation, did they acknowledge there was child sex abuse material on their platform?

Their opening was quite conciliatory in the sense that there was an acknowledgement that inappropriate material, which is a tame word to use-----

To put it mildly.

-----had been available through Grok on the platform of X. I am mindful there is an investigation under way. I do not want to prejudice that in any way. There was a sense of acknowledgement - that is the first word I would use. There was an effort to explain that they had worked furiously over the days of that public outcry to ensure that aspect on X was disabled.

I am sure the Minister of State shares this view - what beggars belief among the public, not just in Ireland but around the world, is that the pendulum has swung so far that even for a couple of days, child sex abuse material or even non-consensual intimate images being put on a platform somehow was juxtaposed into a debate about free speech. How has the pendulum swung so far that there is even a discussion - how could this be permissible for any person legally operating in Ireland or the EU or anywhere where there is and order? How are they even using words like "Inappropriate"? Are they themselves not horrified at the idea of working for a platform - a multi-billion, or trillion platform if you count all of the companies the ultimate owner is involved in - that permitted, however temporarily, child sex abuse material on its platform?

They should be. As the Deputy knows, I disabled my account immediately when I heard this was possible. I have not looked at the platform since. We should be horrified. I made it very clear in no uncertain terms that it was absolutely unacceptable here in Ireland. There was no pushback on that, I have to say, and assurances were given that something like this would not happen again.

How concerned does the Minister of State think that Ireland and capitals across Europe are about other AI platforms? A regulatory framework is in place in the AI Act and, as the Minister of State pointed out, the AI Advisory Council has recommended that Article 5 of that Act should be amended to add harmful practices, such as those we witnessed in this instance. Is that necessary? Is it a step that it might be possible for us to take domestically? When we transpose the AI Act into Irish law and establish the AI office, can we legislate at a domestic level so that no AI platform could produce non-consensual intimate images or child sex abuse imagery? I am amazed I even have to say this, to be honest. Is that something we can legislate for?

I assure the Deputy that, as long as I have this portfolio in this Department, if any steps are required I will take them. As I said, I spent a number of hours with the team and the Attorney General's team to ensure we have all the legislation in place to protect our citizens, including minors. A memo will go to the Government to look at the widening of the prohibited practices and, as the Deputy said, when we transpose the EU AI Act into Irish law, if there is any tightening up or closing of gaps to be done, I will do it.

I acknowledge that the Minister of State said her focus is on the platform and clearly malign actors, potentially people who are on parts of the dark web that only investigating gardaí are familiar with. However, because of how this debate was framed, there may be what might be described as less malign or innocent actors who thought this was somehow amusing, these non-consensual images, or worse, child sex abuse images.

I am going to-----

I will finish on this. We need to have a public campaign, similar to the one that was successfully run about coercive control. I hope that is something the Minister of State will work on with the Department of justice.

I am conscious of time. I call Deputy Gibney.

I hope the Cathaoirleach will give me my extra 90 seconds as well.

I thank the Minister of State for coming before us. I commend her on the strong line she has taken on this, compared with other Government spokespeople. However, it rings hollow. I welcome the European Commission opening an investigation, but three weeks on, image-based sexual abuse continues on this platform. That is just incredible to me and I think it is incredible to most of the public. This is an online safety issue. It is a child protection issue, a data privacy issue, a consumer protection issue, a fundamental rights issue and a criminal issue. The Government has a range of agencies and mechanisms at its disposal to deal with it, many of which it established. Yet, we are still not seeing the action we need. We need to see prosecutions and injunctions. This tool should simply not be available.

Others have focused on the immediate issue, but I want to look at the deeper governance framework because, I, like others, believe this will be like a whac-a-mole; once it has been resolved - if it ever is - we will be faced with more, and a list of potential online harms will not cover it. People will think outside the box, I am afraid, and continue to inflict online harms. People are already being led to suicide by chatbots. People are being radicalised to give birth at home without any medical supervision and dying because of it. People are being pig-butchered out of millions of euro because of online safety issues and the lack of regulation.

The Minister of State and other Government spokespeople have repeatedly cited the AI Act and the Minister of State spoke today about strengthening it. Will she and her Government support the digital omnibus package currently going through the European Union, which will not only gut the AI Act of any potential positive impact, but also roll back on general data protection regulation, GDPR, rights?

Our link to the EU AI board is via the Department of enterprise. With respect to the Minister of State's officials present today and the Department of enterprise and the good work they do, they are not the experts on equality, human rights and fundamental rights we need to represent Irish interests in this. How can the public be assured this approach is the right way to handle this governance? I am hearing a lot that we need to be the best in class on governance and the best regulators, but we do not have the capacity in that Department to do it.

I appreciate Deputy Gibney's passion on this topic. Any time we have had a debate in the Dáil, she has a broader view about mitigating against job losses and what it might do to our labour force, and what she is talking about today, fundamental human rights. I have taken a laser sharp view and focus on what we need to do to protect our citizens.

We cannot say we will block out AI in Ireland.

My two questions were about the digital omnibus and the use of the Department of enterprise as our representatives on the governance framework in the European Union.

Okay, I will finish the point. To say we do not accept or cannot allow AI to be developed in this country-----

That is not what I am saying.

I am trying to answer the Deputy's question as best I can.

I asked two questions. The first was on the digital omnibus.

Yes, we are working on the digital omnibus and I refute the point that simplification in any way guts protection of citizens or humans. In the European view, not just my humble view, we have gone through a decade of over-regulation. I am thinking about enterprise and innovation.

This is my point. Enterprise will always go for deregulation. It is not there to support fundamental rights and it is representing our governance issues.

I am there to support human rights and trying to get the balance right so we ensure we have AI for good, that it promotes innovation, creativity and job creation and our citizens are in no way harmed. By supporting the omnibus, we are all in agreement - Deputy Gibney may not agree with me about this, but anyone in business will tell her-----

It just brings us right back to self-regulation. That is what it does.

Anyone will tell you the regulation we have is over-burdensome and is killing and has killed----

Yet, we have had three weeks of image based sexual abuse and child sexual abuse material, CSAM, being produced. How can the Minister of State say we are over-regulating when this is happening?

We are talking about two different topics, but we have to bring them together to get the balance right.

They are not two different topics.

Human rights and innovation, business and enterprise are two very different things, but we have to find a balanced way of bringing them together that is human-centred, that protects our citizens, but also allows for innovation and creativity.

I thank the Minister of State for being with the committee on such short notice and for the strong stance she has taken in recent weeks. Like my colleague beside me, I welcome the investigation by the European Commission.

What is the timeline for the review of legislation and the gaps? Most of us are aware that it is currently a grey area whether it is illegal to generate non-consensual adult images. We know it is illegal to share them.

The Minister of State said she was looking at options to strengthen reporting. I am concerned that, at the moment, there are teenagers afraid to leave their bedrooms until an image is taken down. How fast is hotline.ie managing to get images taken down? The other option is to report them to the platform and, if it does not take them down to go to Coimisiún na Meán. At what speed are we managing to get images taken down? It is a real impediment. Children are locking themselves at home and refusing to go to school.

We are all focused on Grok this morning, but it just brought this problem into the mainstream. Even if it is conceded there was not intention, we are missing safety-by-design. Grok is not the only problem. It has just brought this into the mainstream. AI chatbots and AI girlfriends are an epidemic at the moment. A study showed that 10% of adults in Ireland had a romantic relationship with an AI girlfriend last year. With a few online searches in AI forums, people will see things like "This AI chick forgets essential stuff" "Most humans are not worth talking to. AI friends are way better" and "I am addicted to my AI girlfriend and it is costing me €200 per month". This is the reality of where we are going. Some of these people are adults, but I wonder about children. What is the cross-government approach to education and awareness right now? The Minister of State is based in the Department of enterprise, but she has said this will be a cross-Government approach.

We are having an urgent review of legislation to see if there are grey areas, but even this week X found another loophole to get around an investigation. We know of things like Candy AI, Replika, character.ai, Ani, Chai AI and Fantasy AI and many of these platforms are not even based here. We can put in all the regulation and legislation and we need to do that, but what urgent actions are being taken on education, awareness and the cross-governmental approach to protect adults who are getting into addictive relationships with AI chatbots and robots that might cost them a lot of money and cause, we do know what kind of damage to real world relationships?

I am especially worried about the children coming up and I would appreciate the Minister of State's thoughts on that.

I thank Deputy Keogh for her very broad contribution. In her analysis of the communications, the public campaign and public awareness, she spoke about young children in their bedrooms speaking to AI bots without realising the harmful nature of some of the content or software they may be using. There is a body of work to be done and I will certainly champion bringing all of the agencies together, including hotline.ie and An Garda Síochána. Deputy Keogh is right that this has to filter through to our schools. I do not know whether it has been her experience but mine certainly has been that schools are reaching out to public representatives and An Garda Síochána. In the constituency of Cavan-Monaghan, An Garda Síochána will run a campaign on the various laws in place and will go into schools to make young people aware of what is illegal.

As legislators we have a job to do to protect these young people before they even get to that point and to protect them from technology that can be harmful. This is where we need the backing of our European colleagues. As I said, I was heartened to hear the comments of Executive Vice-President of the Commission Teresa Ribera this morning. We will also work with Executive Vice-President of the Commission Henna Virkkunen to ensure the Digital Services Act looks into the minute detail of how we can protect our young people against this type of technology, where children are replacing humans with chatbots. It cannot be allowed to persist in this manner. We will need the backing of our European colleagues and the European Commission to do this.

What is the timeline for the review of the legislation gaps? How quickly is hotline.ie getting images taking down?

My understanding is that hotline.ie aims to remove content within 24 hours. A memo will go to Government and I would hope that we deal with the prohibited practice aspect very quickly. I want to be able to introduce it during the European Presidency, with the implementation of the EU AI Act and the establishment of the office. If there are any gaps, we will act on them and introduce legislation as quickly as possible.

I thank the Minister of State for coming before the committee at short notice. I want to talk about her meeting with X. Is it accurate that it told her that Grok, as integrated on the X platform, has been disabled from removing or reducing clothing on individuals worldwide?

Yes on its X platform. Can I-----

That is a quote from the Minister of State's statement. She is certain that is what it told her.

Yes.

Did she check whether this was true?

I did not, because I have disabled my account. I have asked the question and I have been told it has been disabled.

It is not true. I got someone in my office, with my consent, to undress me now. I will spare the Minister of State the image but it is not true and X lied to her. What will the Minister of State do about that?

When I met X, the most dissatisfying part, apart from the fact I had a perception that this involves user manipulation and bad actors, was the attempt to separate Grok from the platform that is X. As we know, Grok is a creation of X as an xAI tool. I made it very clear to X that I do not accept that and-----

To clarify, I checked this and it was done through X. In her statement, the Minister of State draws a distinction between Grok as a stand-alone app and Grok through-----

I do not really draw a distinction, to be clear.

X does, effectively.

Yes.

This is done through X, however. The Minister of State knows what a VPN is. In its statement, X said something different from what it seems to have told the Minister of State. In its statement, it said it now geoblocks the ability of all users to generate images of real people in bikinis, underwear and similar attire via the @Grok account and in Grok on X in those jurisdictions. However, it told the Minister of State it was doing it worldwide. It has not done it worldwide. If people try to access it now in Ireland without using a VPN, they cannot do it through X, but if they use a VPN in Paris, for example, they can still create these images. CBS News did something similar yesterday and found that Grok AI in X still allows users to digitally undress people without consent. Did X mislead the Minister of State?

My understanding from the meeting was that the capability on the platform of X had been disabled.

Was that worldwide?

Within jurisdictions, I would say. I want to be absolutely correct that it is illegal.

In her statement, the Minister of State said X had informed her that the ability to remove or reduce clothing on individuals worldwide was disabled when Grok is accessed through X. That is what she said in her statement. Is that what X told the Minister of State?

It was very clear to me. The purpose of the meeting was to see that it would disable it and that the capability would not be there. My concern is for Irish citizens. Yes, we are part of the European Union but at this moment in time I am concerned about Irish citizens.

Obviously, the Minister of State understands that Irish citizens are not protected if this is simply geoblocked.

Correct, yes.

A large percentage of the population under a certain age have a VPN and are able to say they are somewhere else. I am surprised that no one in the Department checked. The Minister of State was told something by X but nobody seems to have checked whether it was true. It was publicly making statements that were different and was suggesting it had been geoblocked. I was confused about exactly what it was and I am a bit surprised by this. What will the Minister of State do about this? She was misinformed and lied to by X. What will she do about that? It is not exactly clear to me whether she is aware of a criminal investigation into X being under way.

My answer to that is that I have met with Coimisiún na Meán three times at this point. The purpose of the meetings was to seek assurances that an investigation was under way and that there is close collaboration with the Commission. We do need this taken at European level to protect all 27 member states and not only Irish citizens. To answer Deputy Murphy's question, yes, an investigation-----

A criminal investigation.

An investigation is under way with Coimisiún na Meán, in collaboration with the European Commission. As Deputy Murphy knows, the European Commission announced yesterday that it will now undertake a formal investigation into X. I cannot say whether it is criminal because that is not my-----

Is that because it is not happening or because the Minister of State does not know?

I do not know.

X misled the Minister of State and told her something that was not true. She said she welcomed the corrective actions and sought assurances that the capability would not be reintroduced. Either it was reintroduced or it was never taken away. What will the Minister of State do now that she has found out that X has misled her?

I will take action.

What action will the Minister of State take?

I will take action, believe you me. I will take action.

I thank the Minister of State for being with us today. I apologise for being late. I was chairing another committee meeting. I commend the Minister of State on her actions when the Grok controversy arose. I have to say she was extremely fast to move and brought a lot of clarity to the scenario. I commend her on that.

One of the difficulties we have seen in this instance has been raised by others. Be it a criminal investigation or otherwise, it is difficult to hold to account a foreign company. It is a foreign technology company in this instance but it could be any type of company. Essentially, the X office on Fenian Street, as far as I am aware, is essentially just sales, marketing and HR. Does there need to be legislation, or is there enough legislation, to tackle international companies that have an office here in Ireland?

To answer the question as directly as I can, I believe the answer is "Yes".

In her opening statement, the Minister of State mentioned that there is the potential to expand the list of prohibited AI practices under the AI Act. Undressing somebody in a photograph is an obvious issue. Has the Minister of State looked at other AI practices that should be prohibited or considered for prohibition?

A lot of Deputy Ó Cearúil's colleagues in the committee room have alluded to what some of the harmful content might look like. If the committee wants to make further suggestions on what prohibited practices should encompass, I would welcome it.

That is fine.

Within the same sentence, the Minister of State mentioned that she was exploring options to strengthen victim reporting and awareness raising. Is this bolstering and improving hotline.ie? Is it ensuring that social media companies and general technology companies have better reporting mechanisms? What are the Minister of State's thoughts on strengthening victim reporting?

It is all of that. The experience of any of us around this room would be that reporting is one aspect of it. What community standards are within these tech companies? Is the bar high enough? Does the bar need to be higher in terms of the responsiveness and reactiveness in taking down harmful content? It has to be an all-agency approach in terms of heightening awareness to protect our citizens, including Coimisiún na Meán, hotline.ie, An Garda Síochána and the cybercrime unit within that.

To Deputy's Keogh's point about education, I am mindful that we have Departments of education and further education. It will take an all-of-government approach.

Just to inform colleagues - they are probably all well aware - the Taoiseach convened a meeting on foot of this in recent weeks. In attendance were Coimisiún na Meán, the Department of justice, the Department of communications and me. It is about having an all-of-government approach and response.

This is all positive and there is action being taken. However, when a victim is required to report, we need to know that action is being taken. It is all well and good reporting, and it can be quite difficult for people themselves to report. Anonymised data on the number of actions that are being taken by the Garda or the courts against people who are misusing AI or taking advantage of people by using their images and so on would be very beneficial. If we want to improve reporting, we need to ensure that there are actions that come out of it.

It now falls to me. I knew the Minister of State as Cathaoirleach of the media committee in the last term where we served together and I know she is very genuine, but she recalls our experience when we dealt with the Online Safety and Media Regulation Act during that period and our interaction with X. I agree with colleagues that there is a real question about somebody's moral compass if they allow the creation and dissemination of child sexual abuse material. If you are a company that facilitates it, I do not view it as acceptable in any way. I know we have focused a lot on child sex abuse material but there are adult victims here as well. I know the Minister of State attended when Gráinne Seogie presented her testimony to the media committee last week. If you are an adult who is a victim of this, it is pretty horrific.

Obviously, there are a number of things for the Government to consider at this stage and we need to take a strong line. I will make a number of suggestions and ask if the Minister of State agrees with me on these. Obviously, we are looking at addressing platforms such as Dzine. In Germany, I am very conscious that they take Holocaust denial very seriously. If a platform does not address it within a 24-hour period, action can be then taken against the platform. In my view, if a platform does not address the question of child sexual abuse material or some similar material against adults within a quick period, it should be treated as a publisher in the same way as any other publisher would be treated. We need to move in that direction.

The Minister of State will know that, during the online safety and media regulation debates, we had a discussion about the question of making directors individually responsible. Where a company knowingly and wilfully - I am talking by design - allows programmes with this capability to be deployed, it cannot just be about fines. It has to be about individual responsibility. The Minister of State will be aware that I have introduced the Protection of Voice and Image Bill, which is about protecting our digital identities - our voice, image and likeness - and I ask her to look to progress that as part of her response.

I have put down a series of parliamentary questions and I can confirm that no Department is currently advertising on X, which is welcome. By the way, the Department of enterprise was the last one to stop doing it. I hope that the Minister of State would put out a call that no State agency or, indeed, anyone with any sense of corporate social responsibility in Ireland would look to advertise on X.

Those are constructive suggestions I would like to make to the Minister of State. Does she agree with my views on those?

I thank the Cathaoirleach and all of his colleagues for the conversation and suggestions today. I am very open to suggestions because my main goal here is to protect our citizens.

I am aware. I will allow the Minister of State closing remarks, but she might address the specifics.

To answer the Cathaoirleach's question, we absolutely need to move in the vein of these being publishers. I feel very strongly about that. I will be seeking to see whether we can do that.

In terms of the Bill the Cathaoirleach raised, under the AI Act, deepfakes will be watermarked so that we can understand and people can be clear on what is fake and what is real. That is all part of the-----

With respect to the Minister of State, watermarks will not go far enough. This is about protecting someone's digital identity against malign use.

The clearest way to answer any of the propositions the Cathaoirleach made today is for me to go away, work with my officials and see what we can introduce under the AI Act.

I appreciate that. The bottom line is individual liability. I will be straight: I do not trust X. We dealt with it. The Minister of State will recall how, when she was Cathaoirleach of the media committee, it regularly had to be dragged before the committee. At least with the other platforms, whatever the criticisms were, they came before the committees and they willingly engaged. I just do not trust X. My final question is whether the Minister of State trusts X.

Not at the moment, particularly after what Deputy Murphy announced here this morning.

I thank the Minister of State. I am conscious she wanted to deliver some concluding remarks to the committee.

I have a very long speech but I will not subject colleagues to that. I just want to thank the committee. It has been a difficult couple of weeks to try to really get a handle on this. I assure the committee members and all of the people they represent that I want to get a handle on this because I want to protect our citizens. I want to protect our children. I want to try to get the balance right in the sense that we have an ecosystem that allows businesses and creativity to flourish while getting the balance right.

On foot of suggestions from the Minister of State and colleagues, I suggest we look to invite the Data Protection Commission and Coimisiún na Meán to, if necessary, a special meeting of the committee, if colleagues are agreeable. We will liaise with the clerk and so on to try to work out times.

I thank the Minister of State on behalf of colleagues. They very much appreciate her work and the work of her officials on this. There were other aspects of her brief on which we had hoped to engage but I appreciate it was at short notice.

If the Cathaoirleach could indulge me for one second, I thank my officials. As the Cathaoirleach knows, we came at short notice and they had to change plans and everything else. That is not the issue. They have worked with might and main to try to ensure I am informed as I possibly can be and to ensure the legislation is there. We will do anything that needs to be done to ensure enforcement and that we have the capability for enforcement. I thank the departmental officials who are here with me today and the ones who are working behind the scenes.

I thank the Minister of State.