Our AI Flow Talks podcast series shares real-life stories from the world of AI to make AI transformation easier for organizations. This episode delves into European AI sovereignty, technological dependence, and the rapidly changing operating environment. The guest on this episode is Mikko Alasaarela, who discusses these topics with Tomi Leppälahti, AI Director at Fluentia. Mikko is the founder of the Agentics Finland community and Agion, who works at the forefront of the AI revolution and agent-based automation.
The episode begins by discussing Europe's position in relation to the United States and China. According to Alasaarela, Europe has drifted into a vassal position, producing cheap energy and purchasing ready-made artificial intelligence products from elsewhere. As a result, the value chain is flowing out of Europe.
Dependence on US models, cloud services, and component manufacturers – such as Nvidia –combined with geopolitical uncertainty creates a situation in which Europe is vulnerable to the decisions and technological restrictions of other actors. Therefore, alternatives are needed that support both security and competitiveness.
The discussion examines how quickly competition between foundation models is developing and how technically easy it is for companies to switch models today. The challenge is whether organizational processes can keep pace with the constant change.
Interchangeability between models is increasing, but many companies still lack automated testing and quality assurance that would enable them to systematically evaluate the quality and functionality of models. This increases the risk of relying too heavily on a single model – a situation where technical replacement would be easy in itself, but the processes, controls, and metrics cannot withstand a rapid cycle.
Rapid technological development requires organizations to implement new types of process changes so that they can utilize model variation safely and efficiently. At the same time, the discussion notes that agent-based processes have increased productivity many times over in just one year.
One of the key topics of the episode is agent architecture. Future value will not be created by the performance of individual models, but rather by how companies are able to manage their data and build functional agent processes on top of it. The effectiveness of agents is based on their access to up-to-date and contextually enriched information.
In practice, data can be divided into three levels: general data available on the internet, industry-specific information, and a company's own internal data. Internet data is rapidly becoming less reliable, as an increasing proportion of it is generated by artificial intelligence. Industry data, on the other hand, remains an underutilized resource in Europe. A company's internal data – documents, discussions, and tacit knowledge – forms the most critical layer, but is often scattered across different systems. Data and context management built on this layer could become Europe's most significant competitive advantage.
According to Alasaarela, Europe will not be able to compete in GPU production or frontier model training in the short term, but these are not the only ways to succeed. Europe can gain a competitive edge by building world-class infrastructure for the controlled and secure deployment of artificial intelligence.
European operators already have strong expertise in security, transparency, and regulatory predictability. When these qualities are translated into a functional code rather than bureaucracy – for example, into predefined management models for agents – the result is something that may not be found elsewhere: reliability that both European and international companies can rely on.
At the same time, Europe's long-established industrial giants can be elevated to a new level of productivity through agent-based processes. Once digital data sovereignty is in place and European inference and data platforms are developed, the market will be open to new players who can build solutions without being dependent on individual American technology giants.
The discussion highlights that artificial intelligence is not a traditional tool, but rather a whole new way of thinking about processes. Hourly pricing loses its meaning when productivity multiplies. Companies must recognize the value of their own data and learn to mobilize tacit knowledge for agents. Fast-moving and agile players will benefit the most – those who continue with the old model will inevitably fall behind.
Towards the end of the episode, the idea emerges that Europe has the opportunity to turn its values into a competitive advantage. Trust, openness, and responsibility are factors that are in growing demand around the world. The AI revolution can act as a catalyst that will raise European companies to a new level of competitiveness – provided that the change is embraced in a timely and bold manner.
The episode is also available on Spotify
