• The UK’s Competition and Markets Authority proposed a new post‑Brexit tech licensing regime to replace the old EU‑era framework. Reuters

  • Media commentary criticised the UK’s digital ID push, warning that lax oversight by the Information Commissioner risks turning the programme into state surveillance. The Times

  • Senators Hawley and Blumenthal drafted a U.S. bill requiring advanced AI systems to undergo a formal risk evaluation before deployment. Axios


Regulation

United Kingdom / UK tech licensing

  • The CMA wants to retire the EU’s TTBER (Technology Transfer Block Exemption Regulation) and replace it with a 12‑year UK licensing regime focused on clearer rules around patents, software, database rights and competition. Reuters

  • In parallel, the UK is preparing a Cyber Security and Resilience Bill to update its core NIS regulations. This could intersect with AI systems in critical infrastructure. Bird & Bird

Europe / AI Act & mental health

  • The EU AI Act (Regulation (EU) 2024/1689) continues its phased implementation, blending risk‑based rules and interoperability with broader European digital policy. Digital Strategy EU+1

  • Policymakers now debate stronger safeguards around AI’s impact on adolescent mental health. Recent attention focuses on whether the Act’s provisions against “manipulative behaviour” can sufficiently regulate chatbots interacting with vulnerable users. Tech Policy Press

Italy — First domestic AI law

  • Italy passed Law 132/2025 on 17 September, becoming the first EU member to enact a national AI law. Squire Patton Boggs+2Technology's Legal Edge+2

  • The law covers AI deployment across sectors including health, judiciary, public administration, security, and employment. It is scheduled to enter into force on 10 October 2025. Linklaters+1


Cases

  • The English High Court in Ayinde v London Borough of Haringey and Al‑Haroun v Qatar National Bank highlighted risks when parties use AI in litigation settings. The court cautioned that AI tools must respect procedural rules and the duty to the court. White & Case


Academia

  • Red Teaming AI Policy: A Taxonomy of Avoision and the EU AI Act introduces a classification of how firms might strategically skirt or stretch compliance while staying formally within the law. arXiv

  • Scoring the European Citizen in the AI Era analyses how the AI Act’s ban on social scoring may meaningfully curb opaque evaluation systems rather than being symbolic. arXiv

  • The Illusion of Rights based AI Regulation challenges the idea that the EU’s AI regime is grounded in doctrine of rights, arguing instead it is a control instrument adapted to European institutions. arXiv


Business

  • Firms with cross‑jurisdictional operations face significantly more complexity: U.S. state laws (like SB 53) now join the mosaic of regulation alongside the AI Act. Offit Kurman+1

  • In Europe, some voices warn that overzealous rules could stunt innovation — echoing concerns from industry that regulation should be “lean and effective” not burdensome. Reuters


Adoption of AI

  • Companies are already mapping “regulation risk” as a factor in deployment strategies, balancing speed with legal exposure. Offit Kurman+1

  • On user side, public concern over mental‑health effects is pushing some platforms to adopt safety mitigations even prior to legal obligation. Tech Policy Press

  • In the legal domain, practitioners are becoming more cautious about using AI in drafting or research, mindful of court limits and ethical boundaries. White & Case


Takeaway

The spotlight is shifting from rule‑making to rule‑enforcement. The UK is again rethinking its licensing toolkit, and Europe’s AI Act is being stress‑tested over subtle harms like emotional manipulation. For AI governance, the next frontier is remediation: what happens when systems go wrong.

Sources: Reuters, The Times, Axios, Offit Kurman, IR Global, Bird & Bird, Digital Strategy EU+1, Tech Policy Press, Linklaters, Technology's Legal Edge, White & Case, arXiv, arXiv, arXiv