Online Safety (Recommender Algorithms) Bill 2026: Second Stage [Private Members]
Dail Éireann — 2026-03-04
============================================================

Paul Murphy (PBP), Dublin South-West

I move: "That the Bill be now read a Second Time."
The choice we have in this debate and in the vote tonight is whether the profits of social media companies or the health and safety of children are going to come first. That is the choice. The Taoiseach, Micheál Martin, himself said that the damage caused to young people by social media is the public health challenge of our time. The Minister, Deputy McEntee, when asked if we should turn off the algorithms, said we should. Simon Harris, the Tánaiste, said that some things are more important than money when speaking about the need to regulate big tech in the context of the revelations about the mass production of child sexual abuse imagery and imagery of sexual abuse of women by these corporations for profit. Big words; great, but now we have a Bill that implements the recommendation of the AI committee. It implements, to the letter, what the AI committee is calling for. We have a Bill that recommends what is called for by the Irish Council for Civil Liberties, the National Women's Council of Ireland, the National Youth Council of Ireland and all of the experts who have looked at this area, which is to turn off the toxic social media algorithms that are force-feeding toxic content to our young people.
What does the Government say? In its amendment, the Government declines to give the Bill a Second Reading on the grounds the issues the Bill seeks to address are already being addressed by Coimisiún na Meán through its implementation of Ireland's online safety framework. The Government does not quite have the neck to just vote it down because it knows this is popular and is being supported by backbench TDs. Deputy Brendan Smith explicitly called for the measure contained in this Bill about a year and a half ago. It will be interesting to see how he votes tonight. Instead, the Government tries to pretend to people that it is dealing with this issue. It is not dealing with this issue and the reason we know it is not dealing with this issue is that the online safety code, to which it refers, was explicitly amended to take this out. This was in the first draft. It was precisely what is contained in this Bill, which I will go on to explain in a minute. It was amended out because big tech lobbied and said, "Actually, do you know what? The algorithms are great. They protect people and that's why need to keep them". Then Coimisiún na Meán changed it and took it out. Big tech lobbying won out over the safety of our children. This is the Bill that will deal with, or significantly deal with, the issue of toxic content. As I said, it is precisely what all the relevant expert groups are calling for.
To explain what the effect of the Bill would be, it is to turn off the recommender algorithm systems for anyone under 18 and turn them off by default for those over 18. Adults can choose to opt in. If the algorithm is explained in a transparent way and the platform can explain what you are signing up for, you can choose to opt in but it is turned off permanently for under-18s until they reach the age of 18. This would have the effect of turning social media back to what it was supposed to be and what it was sold to us all as, namely, a space where we can connect with our friends, build communities and so on, as opposed to, precisely as Micheál Martin says, a thing that is creating a major public health challenge. There is a huge mental health crisis among young people in particular but in society at large, it is driving a process of alienation, all for the profit of these major social media companies.
Let us be clear: this is why the Government is opposing a measure to protect the health and safety of children. It is because these are huge, powerful corporations that have massive political sway and by which the Government is effectively captured. In regard to the amount of money we are talking about, in 2023, Meta generated over $134 billion in revenue, of which advertising revenue comprised $132 billion. The parent company of TikTok made $120 billion in revenue in 2023. YouTube's global advertising revenues were $31.5 billion in 2023. X was on track to bring in $2.5 billion in 2023. To be clear, these companies are interested in maximising their profits, not in creating safe spaces for young people, women or LGBTQ people. They are interested in maximising the length of time that eyeballs are on the screen so they can maximise the amount of advertising they can sell. That is what they are interested in. Their algorithms are simply designed to do that, and the consequence is the force-feeding of toxic content to young people.
I will give an example. There was a study done by the Centre for Countering Digital Hate called YouTube's Anorexia Algorithm: How YouTube Recommends Eating Disorder Videos to Young Girls in the EU. The study looked at what videos YouTube recommends to a 13-year-old Irish girl who first watches one video about eating disorders. She then gets a thousand YouTube video recommendations. Two in three of those videos that she is then fed over the next couple of hours or whatever relate to eating disorders or weight loss. One in three of those videos is a video advocating harmful eating disorders. One in 20 of those videos involved the promotion of self harm or suicide content. Why are they doing that? Meta derives an estimated $2 million a year in revenue directly from a pro-eating disorder bubble on Instagram alone and $230 million from all those who follow this bubble. That revenue includes that derived from underage users. That is from a Fairplay for Kids report from 2022, Designing for Disorder: Instagram's Pro-eating Disorder Bubble.
There was another study done at Dublin City University, DCU, called Recommending Toxicity: the Role of Algorithmic Recommender Functions on YouTube Shorts and TikTok in Promoting Male Supremacist Influencers. The study set up ten blank male-identified accounts. You go into a shop, buy a new, blank phone and say you are a young man between 16 and 18 years old. All of the male-identified accounts were fed masculinist, anti-feminist and other extremist content within 23 minutes on YouTube. To be clear, these accounts did not watch anything or opt in for any of this stuff. They did not even watch the one video referenced in the study in regard to anorexia and pro-eating disorder content being fed to young women. They did not seek it out at all but one of the accounts was fed this toxic misogyny within two minutes.
It is incredible. Since then, we have had the development of so-called AI girlfriend apps, which is a thoroughly misleading term for apps with taglines like "Build your own AI slut". These apps incite boys to create child sexual abuse imagery by suggesting they upload real photos of people they know from their school or wherever. Within minutes, the app can then create extreme, degrading, deepfake pornography of girls in their class in school.
To go back to the study from DCU, the vast majority of the content being recommended over two to three hours viewing was toxic. It was 76% on TikTok and 78% on YouTube Shorts. By the far the most prevalent feature of this content was Andrew Tate. The other toxic categories were the reactionary right and conspiracy, which accounted for 13.6% of recommended content on TikTok and 5.2% of recommended content on YouTube Shorts. Much of this was anti-transgender content. The report recommended that "recommender algorithms should be turned off by default". The professor involved in the study said, "The social media companies must come under increased pressure from the government to prioritise the safety and wellbeing of young people over profit." This is precisely the choice.
I saw Deputy Ken O'Flynn from Independent Ireland was opposing this Bill and taking the side of the big tech companies over children on the spurious grounds that he is opposed to censorship. To be clear, there is not one iota, none whatsoever, of censorship involved in this legislation. This Bill will have no impact on the ability of people to create whatever content they want and to put it on social media. It simply prevents and stops the force-feeding of this toxic material to young people and enables the rest of us to opt out of it or choose not to opt in to have this material force-fed to us.
This is simple. It really is. Niamh McDonald from the Hope & Courage Collective made the point at a press conference we had that we need to treat big tech as the big tobacco of our day and take action now to prevent even more widespread harms. We have to turn off the torrent of toxic content at its source, the big tech companies and the billionaire psychopaths who own them, instead of taking the Government's approach, as it is pointing to with reference to the online safety code, of putting all the responsibility on individual users and on parents, who are completely overwhelmed by all of this. I appeal to the Government at this stage to withdraw its amendment and pass the Bill. If it has detailed points to make, fine, let us do that in committee, but let us take the side of children and their safety and health and not the side of big tech companies and their profits.

---

Source: Houses of the Oireachtas. Licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). The Official Report is revised after first publication; the fetch timestamp below identifies the version quoted.

Record URI: https://data.oireachtas.ie/akn/ie/debateRecord/dail/2026-03-04/debate/main
Retrieved: 2026-08-14T04:54:43+00:00
Sitting date: 2026-03-04