On May 13, 2026 the Oireachtas Joint Committee on Arts, Media, Communications, Culture and Sport published its report on the regulation of online platforms — and quietly did something that most European legislatures have been unwilling to do this cycle. It declined to recommend a blanket under-16 social media ban. Committee chair Alan Kelly TD, a Labour politician with no obvious incentive to look soft on Big Tech, told reporters that no expert witness across the committee's hearings had recommended one, and that any such ban would be evaded "fairly quickly" once a child reaches for a VPN. The Committee instead issued ten recommendations centred on platform design — chief among them, mandatory deactivation of recommender algorithms for children, default-off recommenders for adults, and a ban on infinite scroll, autoplay, and re-engagement notifications.
The Steelman for a Ban
The case for an age cut-off is not frivolous. Children's mental-health clinicians, the UK's Molly Russell inquest, and now Australia's federal government all start from the same premise: algorithmically optimised feeds reliably push vulnerable minors toward content their developing brains are poorly equipped to resist, and existing safety codes — including Ireland's own Online Safety Code, adopted by Coimisiún na Meán on October 21, 2024 across nine designated video-sharing platforms — have so far failed to address recommender systems at all. A hard age line is at least legible: parents understand it, schools can enforce it, and platforms cannot litigate around it the way they litigate around "reasonable steps." That is the strongest version of the argument, and any Irish policymaker who waves it away is not serious.
Why the Committee Said No Anyway
The Committee's answer was not that the harm is overstated. It was that the proposed cure does not work. Kelly framed it bluntly to RTÉ: "technologically you can get around this very quickly, and we're seeing that in Australia already." Australia's Online Safety Amendment (Social Media Minimum Age) Act 2024 took effect on December 10, 2025, with fines of up to AUD 49.5 million per breach. Early data is grim for the ban's advocates: surveys of 14- and 15-year-olds inside the covered cohort suggest a substantial majority are still on Instagram, TikTok, Snapchat, YouTube and the rest. Polling cited by Australia's Information Age — itself drawing on the Australian Computer Society's own analysis of eSafety Commissioner guidance — shows roughly 30% of evading teens use VPNs, with the remainder simply lying on age prompts or using a parent's account.
The Committee was also alive to a worse second-order effect. Australia's eSafety Commissioner has had to instruct platforms to "try to stop under-16s from using VPNs," which in practice nudges children toward free, ad-supported VPN apps whose data-harvesting behaviour is often worse than the platforms they are routing around. Banning an entire cohort from regulated platforms does not delete the demand; it redirects it to less-regulated rails — Telegram, Discord servers, Coverstar, Rednote — where Irish or EU regulators have neither leverage nor visibility.
What the Committee Recommended Instead
The substantive ask is design regulation. The committee's press release calls for legislation requiring platforms to disable recommender algorithms entirely for child accounts and by default for adults — meaning a user would have to actively opt in to algorithmic curation. Infinite scroll, autoplay and re-engagement notifications would be prohibited outright. Crucially, the report also asks for a "privacy-preserving, risk-based age classification system" rather than the document-uploading, biometric age-verification regimes being trialled in the UK and France.
This tracks a bill already on the Oireachtas order paper. The Online Safety (Recommender Algorithms) Bill 2026, a People Before Profit private member's bill, would amend the Online Safety and Media Regulation Act 2022 to do almost exactly what the committee now endorses. The Government opposed its second reading in March on the grounds that Coimisiún na Meán was already on the case — a position the committee's report has now visibly undercut.
The Pro-Innovation Verdict
For anyone instinctively wary of regulation, the algorithm-controls approach is not a free hit. Default-off recommenders impose real product costs and risk freezing in today's interface conventions for tomorrow's discovery problems. There is a non-trivial speech concern when the state legislates which UX patterns a platform may ship to consenting adults — "infinite scroll" is a layout decision, not a category of harm, and the same code path that delivers doomscrolling also delivers the Wikipedia rabbit hole. The Committee's text should be drafted narrowly enough that an adult user can opt back into an algorithmic feed without a court order, and a startup can ship a video product without first hiring a Dublin compliance team.
But compared with the alternative, this is the more honest policy. A ban that 60–73% of the target cohort ignores within six months is not regulation; it is theatre that erodes the legitimacy of the broader online-safety project. Ireland's committee — which hosts the European headquarters of most of the platforms in question and therefore has more at stake in getting this right than almost any other legislature — has just told Brussels, London and Canberra that the emperor's age-gate has no clothes. That is a useful service, and worth defending against the inevitable political pressure to do something more visible.