Biometric Update reported that Ofcom opened an investigation on 16 July into TikTok Information Technologies UK Limited over its use of 'age inference', a method that estimates a user's age from account activity rather than verifying it. The regulator's Age Assurance report, published the same day, suggested such models may have failed to identify a significant proportion of children, and Ofcom said inference is not listed in its guidance as capable of being highly effective, with an update due in October 2026.
The Irish Times reported that a group of Meta employees has sued the company in California, alleging that AI-driven performance assessment and monitoring systems selected staff for redundancy and disproportionately tagged workers who had taken maternity, medical or other protected leave. The plaintiffs are seeking a preliminary ruling to halt the layoffs while pursuing compensation and reinstatement, and Meta maintains that the decisions were made by human managers rather than by the systems alone.
The Guardian reported on 18 July that incoming Prime Minister Andy Burnham is planning to abolish the Department for Science, Innovation and Technology and fold its functions into the Business department. Politicians, officials and technology industry figures criticised the proposal on the ground that dismantling the department would disrupt the UK's AI and technology agenda at a point when implementation work is under way.
Regulation
The European Commission published guidelines on 20 July 2026 on the transparency obligations for providers and deployers of certain AI systems under Article 50 of the AI Act, which begin to apply on 2 August 2026. The guidelines define the scope of the duties, clarifying which providers must design systems that tell users they are interacting with AI and apply machine-readable marks to generated content, and which deployers must disclose deepfakes and AI-generated publications on matters of public interest issued without human review or editorial control.
Cases
Via the order in Bartz v Anthropic PBC, it can be seen that Judge Araceli Martínez-Olguín of the United States District Court for the Northern District of California granted final approval of the class action settlement on 20 July 2026, entered judgment and dismissed the action with prejudice. The order confirms a non-reversionary fund of 1.5 billion dollars across a Works List of 482,460 registered books downloaded from Library Genesis and Pirate Library Mirror, an estimated per-work payment of about 3,000 dollars, and a claims rate of 91.3 per cent. It also overrules all objections, records that the release does not cover output claims or future conduct, and departs from the percentage of recovery method in favour of a lodestar with a 3.75 multiplier, awarding 101,561,111 dollars in fees rather than the 187,500,000 dollars sought.
Academia
arXiv hosts 'Traccia: An OpenTelemetry-Based Governance Platform for AI Systems', a July 2026 preprint on the gap between the transparency and accountability standards required by frameworks such as the AI Act and what current tooling delivers. The paper argues that fragmented evaluation and monitoring platforms leave agentic architectures exposed to alignment drift and unauthorised shadow deployments, and proposes a layered governance stack combining telemetry, semantic guardrail assessment and execution lineage in a hashed trace ledger.
arXiv hosts 'Overview of Risk Assessment and Management for Intelligent Systems under the AI Act and Beyond' by Roberto Daza and colleagues, submitted on 2 July 2026. The paper surveys the regulatory landscape driving systematic AI risk assessment, characterises the spectrum of identified risks from technical failure to ethical and social impact, and reviews the general frameworks proposed for assessing them while identifying methodological gaps.
Events
The California Lawyers Association will hold Privacy + AI Lab on 17 September 2026 at the Faculty Club, UC Berkeley, with the Berkeley Center for Law and Technology. The full-day workshop-style programme covers disputes, risk assessments, governance and cybersecurity for senior privacy and AI practitioners, with attendance limited.
Legal Business will hold the AI Summit 2026 on 18 November 2026 at The Minster Building in London. The programme covers compliance, ethics, intellectual property and disputes, with panels on navigating global AI regulations and enforcement trends and on building responsible and trustworthy AI systems..
Sources: Biometric Update, The Irish Times, The Guardian, European Commission, Courtlistener, arXiv, California Lawyers Association, Legal Business