AI and data: the keys to innovation in Europe

AI is one of the key drivers of innovation and is therefore crucial for our economy and prosperity. Read the recently published article on this topic by Mediaplanet in Trends.

Share Copied Share on LinkedIn Share on X Share on Facebook

Prefer to read the original article in Dutch by Sylvie Gheysen (Mediaplanet, Trends)? Here it is!

In contrast to the US and China, data sharing in Belgium – and Europe in general – is subject to strict regulations. Frank Hamerlinck, CEO of FTI, Flanders Technology & Innovation, explains the do’s and don’ts within a controlled data environment.

Is Belgium a leader in innovation, or are we lagging behind other countries?

“When we look at the number of startups per capita, we’re doing quite well. More and more unicorns (startups valued at over 1 billion dollars) are also emerging. However, Belgium and Europe are lagging behind the US and China when it comes to AI. Our innovation climate is decent, but the bar must be raised. For promising companies, it’s particularly difficult to scale quickly and secure sufficient funding. Those who aim to become significant players often have to rely on American investors.”

Data is often called the new gold, the raw material of the future.

“Yet, this promise remains largely unfulfilled in Europe. There are few examples of companies that have successfully monetized their data. Large corporations do use user data to support their commercial activities, but the broader societal impact remains limited. The real question is: how can a country as a whole benefit from data? By using it in scientific research, the energy transition, mobility challenges—you name it.”

Frank Hamerlinck: "It is extremely difficult for promising companies to scale quickly and find sufficient resources to do so." ©FTI / Dries Luyten

Is data overly restricted here?

“Yes, the data used to train AI models often comes directly from the US, where it is released with virtually no restrictions. These models can make excellent predictions in an American context but perform less well in a European setting. Ideally, European models should be trained with European data, but complex regulations often prevent this. We aim to break through these barriers: in certain cases, and within the framework of existing laws, data should be made accessible to research institutions, startups, and similar organizations.

In the past, there was a saying: ‘America innovates, China imitates, and Europe regulates.’ While this no longer applies to China, it still does to Europe. By hoarding data, we won’t win the AI race or close the gap. Geopolitically, this is crucial: sectors like the economy, defense, and others will increasingly be driven by AI in the future. To address this, we must avoid a brain drain. Talented individuals who want to develop innovative models often face extremely limited datasets here.”

“That’s unfortunate because a PhD student—educated at society’s expense—who then chooses to work for an American employer represents a significant loss. Job losses due to new technology are often a concern, but innovation almost always creates new industries and jobs in adjacent areas. This time will be no different: data and AI service companies will emerge, and data centers will need to be built. However, if we fall behind, this segment will quickly be dominated by American companies. In California, for example, self-driving cars are already commonplace, while they’re not yet allowed here. It’s likely that, before long, this market will be dominated by players with years of head start. Europe needs to stay vigilant.”

What should AI-specific legislation include to prevent misuse?

“I think there’s a growing realization in politics that we need more transparency around the conditions for data sharing. A good example is the recent change to the prohibition on using medical data for secondary purposes, such as scientific research. This has been adjusted: without explicit consent, patient data—once anonymized—can now be used for research. We’re at the start of a trend to make data more accessible and remove the stigma around sharing it. This is often referred to as data altruism: sharing data for societal purposes. Not only can data help prevent or solve problems, but those who collect or analyze it should also have the right to fair compensation. The good news is that current legislation already offers many possibilities. However, it could be made more specific. Right now, whether data is shared often depends on the personal interpretation of a data protection officer. This process needs to be more transparent and consistent.”

Dystopian visions: AI taking over our lives.

“An AI system trains on historical data and is therefore good at repetitive tasks, summarizing information, and recognizing patterns. GenAI is already capable of creating, but it’s still based on existing ideas. It cannot invent something entirely new or anticipate unexpected situations. The fact that such AI models are fed with data we don’t control highlights the need for regulations to shape our AI industry. We often use tools like Gmail or Google Docs, meaning that all emails, documents, spreadsheets, etc., are effectively in American hands—what is often referred to as ‘data colonization.’ How can we retain and share our data, both in the public and private sectors, in a sovereign manner?

OpenAI—one of the leading AI companies in the US—is currently conducting a funding round of 300 billion dollars. Those figures are staggering. Europe shouldn’t focus on adding more rules but should think about how to close the gap. For example, by investing in secure multiparty computing: instead of pooling all data in one large repository (as is done in the US and China), this technology allows data to be used collaboratively across different owners through a virtual layer. The data remains available but under greater control. Europe still has significant opportunities to develop such technologies.”

FTI advocates for further data sharing and more collaboration among diverse stakeholders.

“We want to bring together as many players as possible from the so-called ecosystem—government, startups, tech specialists, research institutions, and venture builders—in an alliance where data can be shared under controlled conditions. This leads to innovative initiatives. For instance, FTI, in collaboration with Golazo Energy and other partners, has launched the Elli initiative to detect burnout early using data from wearables, surveys, employers, and other sources. This initiative is continually enriched with innovations that add value, such as a company measuring glucose levels.

In the long run, Elli is a win-win-win: socially (lower healthcare costs), personally (improved well-being), and economically (increased productivity). That’s our mission: to initiate and support as many projects as possible to continue driving innovation. Because data is undoubtedly the future.”