ComputerWeekly

Former Italian PM centre stage as Google champions EU sovereignty


Europe can no longer be “competitive, sovereign and small” at the same time, Enrico Letta told an audience in Brussels this week. The former Italian prime minister – and now dean of politics, economics, and global affairs at IE University in Madrid – was speaking at an event hosted by Google, rather than the European Parliament or a national capital.

“In the past, the discussion about sovereignty, dimension and competitiveness was a very simple discussion,” he said. “In the past, you were able to be competitive small, and sovereign at the same time. Today, you can’t be competitive, sovereign and small. You need to be integrated in something bigger, and this is what we need to do, to scale up.” 

Letta spoke at Google’s Sovereign AI Summit in Brussels this week, where the US hyperscaler argued that European strength and European sovereignty can advance together.

Marianne Janik, who leads Google Cloud’s operations across EMEA North, opened the event with a message that presaged Letta’s. “AI will shape the next decade of Europe’s public service, healthcare and finance, and has to happen under European control,” she said.

Janik said the same four questions surfaced “in conversation after conversation”. Namely: Who can access our data and can we prove to regulators that our controls work? How do we keep AI running reliably? Will choosing more control mean settling for less capable AI? And will the value created with AI stay in Europe?

Her answer was that Google can help organisations with all four of these. “Keep your options open at every layer,” she told delegates, describing a stack that customers can compose from European partners, open models and Google’s own infrastructure.

The company used the summit to put Gemini models into Google Cloud Dedicated, preview confidential external key management and add audit frameworks. These extend the three-tier sovereign portfolio – data boundary, dedicated cloud and distributed cloud – it launched last year. It also launched an accelerator for sovereign AI startups and said more than a dozen software suppliers, including European providers, had validated their products on its dedicated cloud.

Letta, author of the 2024 report Much more than a market, argued that Europe’s problem is fragmentation rather than a lack of ambition. He wants a “28th regime”, a European code of business law that lets firms scale across the bloc without 27 national rulebooks.

“I don’t want Europe being a protectionist fortress,” he said. “I would like to see Europe stronger, more integrated, able to recuperate 25 years of non-integration.”

He framed the debate around what he called the “cost of non-innovation”, borrowing from the 1988 Cecchini report on the cost of non-integration that underpinned the single market. The same method, he argued, should now apply to technology. Namely, every regulation that slows adoption carries a price that rarely appears on the ledger.

He was also clear that the answer is not to shut out US technology. “If you think that now we get total energy independence in one year, total financial independence in one year, total tech independence in one year, I think it’s totally unrealistic,” he said. Trust, he added, “is the most relevant value in our world”, and “mutual trust between Europe and the US is also politically fundamental”.

The politics turns on CADA, the Cloud and AI Development Act, which European institutions are still negotiating. Google’s stage guests warned that a restrictive reading would be expensive. Discussion cited an ECIPE paper that put the migration bill at €80bn across the next decade and warned of an eight-year delay to innovation.

The fault line is whether CADA should favour “made in Europe” providers or stay open to trusted non-European firms that operate through local partners. Google is lobbying hard for the latter, and its guests echoed the argument that exclusion would raise costs without improving security.

Giorgia Albetino, senior director and head of government affairs and public policy for Google Cloud EMEA, put it directly. “I don’t believe that only made-in-Europe will really serve the interests of Europeans,” she said.

The dependency argument ran through the event. Timo Kleiner, director of future networks at the European Commission, said “dependency is a problem” when a single provider fails. “Dependencies, when they are too significant, become risks,” he said, referring to the disruption caused by a global IT outage the previous week.

Marcus Mühling, chief executive of Secure Cloud Germany, put the customer view plainly: “If your clients have to make extreme choices – either sovereignty or innovation – that’s broken.”

The event was not a one-way conversation. Roy Illsley, an analyst at Omdia, read the display as a competitive move. “Google is trying to create a difference between its approach and their rivals,” he said. “They are playing the ‘open’ card, which has been under-used as a contingency or alternate supply.”

Oracle has Alloy, AWS has its European Sovereign Cloud, and Microsoft is “playing catch up”, Illsley argued. Google’s enthusiasm for European integration, on this reading, is a way to stand apart from the pack.

It might appear puzzling that a US hyperscaler would platform explicit European advocacy. But Google has a commercial interest in the outcome. A louder European voice on sovereignty helps shape CADA in a direction that keeps hyperscalers inside the tent, and it positions Google as the partner of choice for governments that want control without giving up frontier AI.

For UK IT leaders, the event matters less as a product launch than as a signal. UK CIOs should be watching how far CADA reaches, since many run workloads in EU regions and face the same questions about who can access data.

Google has brought an interesting pitch – a company subject to US law, arguing for European strategic autonomy. It has made the argument louder than its rivals and attached a product portfolio to it.



Source link