Colao: Europe missed one wave, but AI is the next test
Vittorio Colao, the former chief executive of Vodafone and a former Italian minister for technological innovation and digital transition, used a closed-door press Q&A at Wave by Vento in Turin to deliver a blunt message about Europe’s place in the global technology race. Europe has less than 2 gigawatts of AI computing capacity and needs far more, he said. The figure is a stark reminder of how far the continent trails the United States and China in the infrastructure that powers artificial intelligence.
Colao ran Vodafone from 2008 to 2018, a decade in which he reshaped the telecoms group around mobile, broadband and enterprise services. He later served in Mario Draghi’s government as Italy’s minister for technological innovation and digital transition. That combination of corporate and public experience gives his diagnosis particular weight. He is not a techno-optimist who ignores policy, nor a bureaucrat who ignores markets. His argument is that Europe has the skills, the companies and the savings, but it lacks the speed, the scale and the regulatory simplicity to turn them into global winners.
‘Europe missed the big digital wave. Fine, 20 years,’ Colao said. The line is both an admission and a challenge. Europe missed the consumer internet wave, the cloud wave, the social media wave and much of the mobile platform wave. American and Chinese companies built the dominant platforms, while European firms often excelled in niche engineering, industrial software or telecommunications. Colao’s point is that the next wave—artificial intelligence—is still forming. Europe should not miss it.
He said Europe is starting to put money into computing. The speed is not the right one, but the skills are there. That is a familiar critique. The European Union has launched a series of initiatives, from the European Chips Act to AI gigafactories, to build more high-performance computing and data centre capacity. But the gap remains large. AI models require enormous clusters of specialised chips, reliable energy and fast networks. Europe’s fragmented energy market, permitting rules and capital markets make it harder to build at the pace of American hyperscalers. Colao’s 2-gigawatt estimate is a measure of the challenge: less than 2 GW of AI computing capacity is not enough for a continent that wants to host the next generation of AI companies.
Simplify the first three to five years
The most important thing governments can do, Colao said, is make the first three to five years of a company’s life far simpler. He was speaking in Italian, and his examples were concrete. Italy, Germany and France remain places of administrative and labour-law complications. For a young company, those complications are not abstract. They consume management time, legal fees and cash that should go into product and customers. They also make it harder to hire quickly, restructure when needed and expand across borders.
He gave the example of a digital health company he works for. It was amazingly successful in France and had good success in Germany, but already with a different model. Moving into the UK required another model. Coming to Italy required yet another model. The point is not that European countries are hostile to business. It is that each market has its own rules, its own labour practices, its own tax treatment and its own administrative culture. For a startup, that means duplication. For a scale-up, it means friction. For a global competitor, it means Europe remains a collection of national markets rather than a single home market.
Colao called what has emerged from Europe’s so-called ‘28th regime’ for companies ‘un topolino’—a little mouse. The 28th regime is an idea for an optional EU-wide corporate framework that would sit alongside the 27 national company law regimes. In theory, it would let innovative companies operate under one set of rules across the single market. In practice, Colao said, no member state—Italy included—proposed adding labour, tax and administrative rules. So the project stopped at the simple part: registration. A little mouse, not a tiger.
He also wants public procurement opened to small companies, with a small share of contracts set aside for them. Public procurement is a huge part of European GDP, but it often favours large incumbents that can navigate complex tenders. Setting aside a share for smaller firms could give startups a first customer, a reference and revenue. That is a practical industrial policy that does not require picking winners. It simply widens the funnel.
Colao is sceptical of governments acting as venture capitalists. A state fund of funds is fine, he said. So is public money for strategic projects such as chips and energy. Those are industrial policy, not venture capital. The distinction matters. Venture capital is about taking equity risk in young companies, accepting many failures and relying on market signals. Industrial policy is about building infrastructure, securing supply chains and correcting market failures. When governments blur the two, they often end up with politically protected companies that cannot scale globally.
A single capital market with depth
Asked how Europe can keep companies from listing in the United States, Colao said it needs a real single capital market, or savings and investment union, with depth. Euronext and the German and Swiss exchanges together have nothing compared with the American market, he said. That is not a criticism of those exchanges. It is a statement about scale. The US has deep pension funds, large venture capital firms, a vast pool of institutional capital and a culture of risk-taking. Europe has savings, but they are often parked in bank deposits or government bonds. The result is a funding gap for scale-ups.
‘We can write yet another report. But it’s time to start doing,’ Colao said. Europe has no shortage of reports on capital markets union, innovation, competitiveness and the single market. The Draghi report on competitiveness and the Letta report on the single market both diagnosed the problem. The difficulty is implementation. Each member state agrees in principle and then asks for something that slows the system down. A single capital market requires common supervision, insolvency rules, tax treatment and pension fund frameworks. Those are exactly the areas where national sovereignty is most jealously guarded.
Colao said he is more optimistic about European investors’ appetite for risk than he was five years ago. Two things have to change. First, successful founders must reinvest their money, as they do in Sweden. Second, pension funds must put money into venture. The first is already happening. Europe has produced a generation of founders who have sold companies or taken them public, and some are now angel investors, limited partners and mentors. The second needs rule changes. Pension funds are often constrained by regulations that limit risky assets. Changing those rules could unlock a wave of capital for European venture and growth funds.
At the same event, Bending Spoons’ Luca Ferrari argued that capital is no longer Europe’s big barrier. That view is increasingly common among European founders. The problem is not only money. It is the ability to scale across fragmented markets, to hire and retain talent, to access deep public markets and to compete with American giants that can deploy capital at a pace European companies struggle to match. Capital matters, but so does market design.
What growth investors look for
Colao was also asked what he looks for in a founder. The single thing, he said, is an obsession with customers and product. A founder focused on customer quality from the first few million in sales is the best sign that a company can scale to 500 million. He also wants clean, detailed revenue figures. That is a classic growth investor’s checklist. Revenue quality, retention, unit economics and customer satisfaction matter more than a glossy pitch. In Europe, where many startups operate in complex B2B markets, customer obsession is often the difference between a niche consultancy and a scalable software business.
The emphasis on clean revenue figures is also a warning. European scale-ups sometimes lack the financial discipline of their American peers. Fragmented reporting, unclear KPIs and optimistic projections can make it harder to raise growth capital. Investors want to see that a company understands its own numbers. That discipline becomes more important as companies expand across borders and add layers of regulation.
AI risk and cyber defence
On AI, Colao said there is some risk, partly because development ran as a race without the safeguards a more traditional company would have added. The point is not that AI is uniquely dangerous. It is that the speed of development has outpaced the institutions that normally manage risk. A pharmaceutical company cannot launch a drug without clinical trials. An airline cannot fly a new plane without certification. AI models, by contrast, have been deployed at global scale with limited external oversight. Because AI works at global scale, the safeguards need rules everyone respects. That means international cooperation, not just European regulation.
Cyber defence needs much closer cooperation, because reaction time matters, he said. He described the UK’s National Cyber Security Centre as ahead of others. Europe did the right thing in creating its cyber security authorities, but they need more resources. The threat landscape is increasingly cross-border. A ransomware attack on a hospital in one country can be traced to servers in another and paid for in cryptocurrency in a third. National authorities can only do so much alone. Intelligence sharing, joint exercises and common standards are essential. The European Union has built institutions such as ENISA and frameworks such as NIS2, but Colao’s point is that resources and speed must match the threat.
Europe is also too self-critical, he said. That may sound surprising from a speaker who began by saying Europe missed 20 years of digital development. But self-criticism can become paralysis. Europe has world-class universities, strong industrial bases, a large single market and a growing startup ecosystem. In July, the EU opened bidding for AI ‘gigafactories’ to add computing capacity. The goal is to create large-scale facilities that can train and run advanced AI models. On Thursday, a data platform chief executive told the same event that Europe’s first trillion-dollar startup may already exist. Whether that is true or not, the remark captures a shift in mood: Europe is no longer only asking what it lacks, but also what it can build.
Source: TNW | Government-policy News