The second political trilogue on the EU's AI Act Omnibus collapsed on 28 April after more than 12 hours of talks. Parliament, Council, and Commission could not agree on the proposed deferral of high-risk AI obligations. A third session is scheduled for 13 May. If the Omnibus is not formally adopted before 2 August 2026, the original AI Act's high-risk provisions take effect on schedule, catching thousands of organisations mid-preparation with no regulatory buffer. Ninety days.
(The Next Web) (IAPP)
Policy & Regulation
DMA Passes First Review — Scope Stays Put
The European Commission published the Digital Markets Act's first mandatory evaluation on 28 April, concluding the regulation is "fit for purpose" and requires no substantial revision. Calls to expand the DMA to cover generative AI services and social media interoperability were rejected as premature; the Commission wants more evidence before legislating. Cloud is where attention now turns: market investigations into Microsoft Azure and Amazon Web Services are expected to produce gatekeeper designation decisions by November 2026, with full conclusions on whether the DMA adequately addresses cloud concentration due by May 2027.
Cybersecurity Package Takes Shape
The European Commission proposed three significant updates to its cybersecurity framework in Q1 2026: a revised Cybersecurity Act, a new Digital Networks Act to modernise EU telecoms rules, and proposed amendments to NIS2 intended to ease compliance burden. The Cyber Resilience Act, which governs connected devices, also received draft enforcement guidance from the Commission; across the board, Brussels is moving from legislation to implementation.
DSA Second Enforcement Wave
The Digital Services Act's second enforcement wave is underway in Q2 2026, extending Very Large Online Platform obligations to a broader set of mid-tier companies, including several US-based services that had not anticipated being caught. These platforms now have six months to comply with the full VLOP rulebook: algorithmic transparency, risk assessments, and researcher data access. For platforms that miscalculated their exposure, the reckoning is arriving faster than expected.
UK Drops AI Copyright Opt-Out
The UK government published its copyright and AI report in March 2026, retreating from its previously preferred opt-out model that would have allowed AI developers to train on copyrighted material unless rights holders explicitly blocked it. No new option has been selected; the government is now weighing licensing requirements, a broad data-mining exception, and no change at all. For UK creative industries and AI developers alike, the uncertainty extends at minimum into 2027.
Capital & Investment
Ineffable Intelligence Sets European Seed Record
London-based Ineffable Intelligence emerged from stealth on 27 April with a $1.1 billion seed round at a $5.1 billion post-money valuation, the largest seed round in European history. Founded by David Silver, former head of reinforcement learning at Google DeepMind, the lab is building a "superlearner" that acquires knowledge and skills without relying on human-generated training data. The round was led by Sequoia and Lightspeed, with NVIDIA, Google, Index Ventures, and DST Global participating alongside the UK Sovereign AI Fund and the British Business Bank. The government co-investing at seed stage is the signal worth reading; it is an attempt to anchor frontier AI capability in the UK before it relocates.
(TechCrunch) (CNBC)
EIC Commits €146.5M to Eight Scale-Ups
The European Innovation Council announced on 27 April that eight European startups will receive equity investments of €10–30 million each under the STEP Scale Up scheme, totalling up to €146.5 million pending successful negotiations. The scheme targets the structural gap between early venture funding and commercial scale, a zone where European companies have historically stalled or relocated. The sectors were not disclosed ahead of individual negotiations, but the EIC's recent pattern favours deep tech, health, and strategic industrial technology.
Pre-Seed Momentum in Northern Europe
Danish VC DFF Ventures closed its third fund at €70 million in April, targeting pre-seed and seed investments across software, deep tech, and digital health. The close is a small but clear signal: LP appetite for early-stage European venture is holding even as mid-stage rounds remain selective and exit timelines stretch. DFF joins a handful of specialist Nordic and Baltic funds that have raised new vehicles in Q1–Q2 2026.
Talent & Workforce
The Frontier Lab Pattern Crystallises
David Silver's departure from Google DeepMind to found Ineffable Intelligence clarifies something about European AI talent. A world-class researcher chose to stay in Europe, but not inside a corporate lab. He built an independent frontier structure, raised at sovereign scale, and attracted a mix of global VC and UK public capital. This is the second major European AI lab this year to follow that pattern. The question for policymakers is not whether Europe can produce frontier researchers; it is whether the conditions exist to keep them here once the money gets serious.
(CNBC)
One to Watch
Cloud Gatekeeper Designations
By November 2026, the European Commission must conclude its DMA gatekeeper investigations into Microsoft Azure and Amazon Web Services. If designated, the hyperscalers would face the full DMA obligations in a cloud context for the first time: interoperability, data portability, no self-preferencing. A parallel investigation is assessing whether the DMA is structurally adequate to address cloud concentration at all, with legislative amendments possible if not. The November deadline is the most consequential test of whether European digital sovereignty is enforceable or aspirational.