Thelonious ICYMI — Edition 11 — 18 June 2026
Welcome to the 11th edition of ICYMI — your weekly briefing on the developments shaping the AI legal and regulatory landscape, curated from Thelonious.
Here’s a snapshot of what’s inside this week:
- European Parliament Approves the “Nudifier” Ban — On June 16, the European Parliament gave final approval to a set of amendments to the EU AI Act, carried as part of the wider Digital Omnibus package. The marquee provision bans AI systems that generate child sexual abuse material or non-consensual intimate imagery — the so-called “nudifier” applications. It is a ban no lawmaker will be seen to oppose. Travelling in the same package, with far less fanfare: postponed compliance deadlines for high-risk AI systems and watermarking obligations, simplified requirements for AI-enabled machinery, extended exemptions for small mid-cap enterprises, and streamlined enforcement. The ban is the headline. The deadline extensions are the policy. When a morally unarguable prohibition arrives bundled with relief on the obligations developers actually lobbied against, read the timeline, not the press release.
- The 2026 State of the Digital Decade Report — The European Commission’s annual assessment of the bloc’s digital transformation landed this week, crediting solid progress on connectivity while flagging significant shortfalls in foundational technologies, computing capacity, cybersecurity, and digital skillsIt is a striking admission from the institution that has written more AI regulation than any other. Europe sets the rules the rest of the world reacts to, yet its own scorecard concedes it is short on the compute and skills needed to build the technology it governs. You can regulate what you do not build — but not indefinitely.
- Canada’s Safe Social Media Act — Bill C-34 proposes a new online-safety framework for Canada, imposing duties on social media and AI chatbot services to protect users — particularly children — from seven defined categories of harmful content. The detail worth marking is the scope. Chatbots are pulled into the same statutory duty of care as social platforms. The regulatory category “social media” is quietly expanding to mean anything that talks back.
- The UK Puts an Age Rating on Artificial Affection — The UK government published a policy statement on June 15 setting out plans to bar social media platforms for under-16s and impose a minimum age of 18 for AI romantic companion chatbots, with similar intimate functionalities restricted for under-18s across other AI services. Britain has, in effect, classified AI companionship as an adult product. The restriction itself is less novel than the official recognition underneath it: that a relationship with a machine is now a thing with a maturity rating.
- Canada’s National New AI Strategy — Canada has published a national strategy to harness AI for economic and social benefit, built on six pillars: protecting citizens, building skills, driving adoption across the economy, establishing sovereign compute infrastructure, scaling domestic AI companies, and forming trusted international partnerships. “Sovereign compute” is the tell. The contest is quietly shifting from who regulates AI to who owns the hardware underneath it.
- Florida v. TikTok — The State of Florida alleges that TikTok admits underage users without parental consent, knowingly designs addictive features that damage the mental health of minors, and misleads parents about the frequency and intensity of harmful content on the platform. Florida frames the conduct as deceptive trade practices and a public nuisance. “Deceptive trade practices” is the predictable claim. “Public nuisance” is the one to read twice — the same legal theory that unwound Big Tobacco and the opioid manufacturers, now pointed squarely at a recommendation algorithm.