European Voices Converge on Reliable AI
Since we published our case for why Europe should bet on reliable AI infrastructure last month, the momentum surrounding reliable AI in Europe has grown.
In November, France, Germany, and the European Commission launched the Frontier AI Initiative. Importantly, the initiative explicitly aims to develop AI systems that are “more reliable for our industries, more aligned with our values, more complementary to humans’ skills.”
This framing suggests France and Germany are pursuing their own vision rather than simply replicating approaches from elsewhere. Reliability and alignment appear as core objectives alongside performance.
Voices across Member States
This week, former MEP Marietje Schaake made a complementary argument in the Financial Times. Writing as a Dutch perspective, she articulates why merely pursuing a hyperscaler model doesn’t serve European industry needs:
“A German car manufacturer doesn’t need a chatbot trained on the entire internet. It benefits from AI systems trained on high-quality engineering data to optimize manufacturing processes. A Dutch hospital needs diagnostic tools that meet medical standards. A French bank needs AI that delivers efficiency while adhering to strict financial regulation.”
Her conclusion points to the opportunity in developing reliable AI systems, something current general-purpose models struggle to deliver.
Europe can fill this gap and forge a new approach.
This message resonates across sectors. Vanessa Cann, Managing Director and Data & AI Innovation Lead at Accenture, highlights that AI systems need to maintain 99% accuracy on task execution to succeed in the market. Alstom emphasizes building trust in AI as central to safe, reliable mobility solutions. And it’s not just industry – according to recent studies, more than 60% of healthcare professionals have expressed hesitation in adopting AI systems due to lack of transparency and fear of data insecurity.
What’s significant is the breadth of this emerging consensus. The Frontier AI Initiative is Franco-German, whilst Schaake brings a Dutch voice. Industry leaders from Accenture to Alstom are making the case. As the Tony Blair Institute recently argued, pursuing this path will make Europe a trustworthy global partner that presents the world with a technology stack built on European values, rule of law, and a commitment to shared prosperity.
From agreement to action
Agreement on direction creates the opportunity for decisive action. Our previous post outlined three priorities: positioning the Frontier AI Initiative around reliability, safety, and security; transforming RAISE into a genuine ARPA-style institute; and ensuring Gigafactories serve a strategic, European vision, rather than simply providing compute. The coming months offer a chance to implement this vision.
The encouraging news is that the conversation has shifted. Six months ago, discussions about European AI strategy felt limited to “how do we compete on scale?” Today, more voices are asking a different question – what can Europe build that others aren’t building?
What we’re watching
Several decisions in the coming months will shape whether this momentum translates into a coherent European approach.
Member states and stakeholders should coordinate around a focused strategy pursuing reliable, safe and secure AI. Franco-German leadership is a strong start. Bringing additional member states into alignment, along with industry, academia, and civil society, will be even stronger. With clarity and unity, Europe can bet on a strategic direction and execute.
RAISE should become an empowered, ARPA-style institute. With substantial funding, a culture for breakthrough innovation, and program managers who can make bold bets, RAISE could rapidly advance European AI research faster than a virtual network which coordinates between partners.
The Frontier AI Initiative should prioritize reliability, safety and security as central research workstreams, and recruit additional member states. This means remaining agile enough to fund ambitious moonshot projects quickly, and attracting visionary researchers and research managers who can set direction.
We’ll be following these developments closely and will share updates as the picture becomes clearer.
