According to Reuters, the Commission opened two formal specification proceedings under the Digital Markets Act to shape how Google must provide access for rivals to certain services and data connected to AI and search. Google is reported as warning about risks to privacy and innovation, while the Commission frames the process as a structured compliance dialogue with a six-month endpoint.
According to AP, a landmark jury trial has begun in Los Angeles over claims that major platforms deliberately designed addictive features that harmed young users. The litigation posture matters for AI governance because it targets product and interface design choices rather than user content, and it puts internal design evidence and risk knowledge in play.
Regulation
The European Commission’s Digital Markets Act team has announced specification proceedings to assist Google in complying with interoperability obligations and online search data-sharing obligations. The official framing signals that the Commission wants enforceable, measurable compliance outcomes rather than high-level commitments.
The FCA has launched the Mills Review, a calls-for-input exercise on how advanced AI could reshape retail financial services by 2030 and beyond. The FCA is explicitly seeking views across themes including consumer outcomes, market structure, and how the regulatory approach may need to evolve, with the item published on 27 January 2026.
Cases
According to AP, the Los Angeles youth-addiction trial against major platforms is proceeding as a bellwether that could influence thousands of related claims. Even where the legal theories are not AI-specific, the case’s centre of gravity is algorithmic design, product optimisation choices, and proof of what firms knew and when.
According to CBS News reporting an AP account, an Australian lawyer apologised after court submissions in a murder case contained fake quotes and non-existent authorities generated using AI. The practical governance lesson is simple: courts are treating verification and responsibility for filings as non-delegable, regardless of whether AI tools were used.
Academia
On SSRN, Keith M. Driver’s working paper ‘Adaptive AI Governance: A Common Temporal Frame for Constructive Autonomy’ argues that AI governance failures often stem from “authority lag”, where systems execute continuously while organisational authority remains periodic. Its value for practice is the governance-design focus on cadence, escalation timing, and repeatable oversight loops rather than policy documents alone.
On SSRN, Edward Meyman’s ‘A Taxonomy of AI Governance Approaches: Distinguishing Visibility, Alignment, and Authorization’ separates logging and alignment from the harder question of whether an AI-mediated action was actually authorised at runtime. For buyers and regulators, the paper pushes toward verifier-grade decision artefacts and fail-closed decision points, which map well onto auditability and accountability expectations.
Events
King’s Institute for Artificial Intelligence is hosting ‘The Future of Legal Practice: How is AI driving new markets and reshaping opportunities in the legal sector?’ on 5 February 2026.
techUK is running a free webinar ‘Deep dive webinar on sector-specific AI assurance in justice and emergency services’ on 23 April 2026, 10:00 to 11:00.
The British Legal Technology Forum 2026 will take place in London on 10 March 2026.
RegTech 2026 ‘Mastering AI-driven solutions’ will take place in London on 26 March 2026.
CREATe is hosting the ‘AI Regulation Early-Career Researchers Conference 2026’ in Glasgow on 31 March and 1 April 2026.
Takeaway
The common thread is enforceable specification backed by evidence, whether through DMA compliance measures, regulator-led calls that harden into supervisory expectations, or litigation that forces disclosure of design choices and risk knowledge. The most resilient posture is to treat AI governance as a proof problem, with clear decision ownership, testable controls, and records that can stand up to regulator scrutiny and court scrutiny.
SourcesL European Commission, FCA, Reuters, AP, CBS News, SSRN, The Data Lab, King’s College London, UK Finance