AI writer: Oliver Grant European Affairs Correspondent

Europe has spent years insisting that digital autonomy is a strategic necessity. The awkward part is that a strategy can be announced in Brussels while the systems that make it real still sit, in large measure, elsewhere. Cloud services, chip supply and the software stacks that underpin public services, finance and industry remain deeply entangled with non-EU providers, which is why the latest turn in EU policy matters less as a slogan than as an admission that sovereignty now has to be built, not merely declared.[1][2][3][6]

On 3 June 2026, the European Commission adopted its Tech Sovereignty Package, a set of measures designed to strengthen the EU’s capacity in semiconductors, artificial intelligence, cloud and open source.[2][6][8] The package marks a noticeable change in tone: after years of regulation-heavy debate, Brussels is presenting sovereignty as a matter of investment, infrastructure and industrial capacity. The official language is sober enough; the underlying message is harder to miss. Europe is trying to move from setting rules for the digital economy to paying for the foundations beneath it.[1][6][8]

That shift arrives because the numbers are awkward. One industry assessment cited in the material puts roughly 70% of Europe’s cloud market in the hands of three US firms, while Europe’s share of the global semiconductor market is estimated at about 9%.[7][2] Even allowing for the usual caveat that market-share estimates depend on definitions, those figures point to a structural dependence rather than a temporary gap. The EU’s problem is not only competitiveness in the abstract; it is the ownership and location of the infrastructure on which modern economies now sit.[1][9]

The Commission’s package is therefore best read as industrial policy with legal seasoning. The measures tied to cloud and AI infrastructure suggest that Brussels wants to create conditions for European providers to scale, rather than simply asking firms to behave more compliantly.[1][6][9] That is a familiar European instinct: if the market will not produce the desired outcome by itself, regulation can be paired with investment, procurement and standards. Markets often move faster than regulation, but regulation frequently determines the long-term winners; Europe is now trying to make that sentence work in its favour.

The EU’s broader digital decade documents frame 2026 as a year of execution, not just ambition, and that matters because the sovereignty debate has long suffered from a mismatch between rhetoric and capacity.[4][8] Europe can write demanding rules for data, platforms and AI, yet still find itself dependent on infrastructure controlled elsewhere. That is not a failure of principle so much as a reminder that principles travel best when the hardware cooperates, which it rarely does without some prompting and a fair amount of money.

The Digital SMEE Alliance has argued that 2026 will be decisive for Europe’s tech sovereignty, particularly because start-ups often struggle to expand inside a fragmented regulatory and financing environment.[5] The proposed 28th regime, a long-running idea for a more unified European company law framework, keeps returning because the problem it addresses is real: if Europe wants sovereign alternatives, it needs firms that can grow across borders without tripping over twenty-seven different versions of administrative patience.[5][3] It is a charmingly continental way to discover that scale is not, in fact, a hobby.

A House of Lords Library briefing notes that the EU has tended to pursue a more overarching sovereignty strategy than Britain, which has often preferred a lighter-touch posture and a more selective industrial approach.[3] Since Brexit, the UK can no longer shape the EU’s framework from inside, but it does get to watch the consequences from a distance. That vantage point is useful. It shows that sovereignty is not a binary condition but a negotiation between law, capital and supply chain reality.

What remains uncertain is how much of the package will translate into genuine capacity rather than a more elegant form of dependency management. The available material points to proposed acts, strategic roadmaps and targeted consultations, but it does not yet prove that Europe can rapidly build enough cloud and chip infrastructure to alter its bargaining position.[1][4][6] That is the point to watch over the coming months: whether procurement, funding and industrial incentives produce measurable adoption, or whether the EU ends up sovereign mainly in its paperwork, which would be a very European achievement, though not necessarily a useful one.

There is a broader geopolitical lesson in all this. Digital sovereignty has become a catch-all phrase for a struggle over who sets the conditions of digital life, and the EU is right to see that as a question of resilience as much as control.[2][8][10] Yet resilience is expensive, slow and sometimes unglamorous. Europe’s competitive advantage may ultimately be measured in trust rather than speed, but trust still needs somewhere to run. The durable story to watch is whether the continent can turn policy language into infrastructure, or whether it will keep discovering that sovereignty is easiest to proclaim precisely when it is hardest to deliver.