Founders' perspectives: The necessity of having European AI professionals.

      Founders’ Takes is a new series offering expert perspectives from tech leaders revolutionizing industries through artificial intelligence. In this edition, Lucas Spreiter, founder of the German startup Venta AI, shares his vision for AI employees.

      Artificial intelligence is on the verge of facilitating the most significant transformation of the century: shifting from human labor to AI labor. In the years ahead, businesses will not just utilize AI as a tool — they will integrate AI as true coworkers, managing critical workflows from start to finish.

      This transition is unavoidable. The pressing question is: whose employees will we be bringing on board? If Europe fails to keep pace with the US and China and does not develop its own AI workforce, we risk outsourcing a considerable portion of our economic value creation — or as we Germans refer to it, Wertschöpfung: the essence of how we generate wealth.

      State of AI

      The breakthrough occurred with OpenAI’s ChatGPT in 2022. It demonstrated that knowledge work could be automated at scale: tasks such as research, writing, coding, and analysis suddenly became feasible through software rather than human effort. The arrival of AI agents extends these capabilities even further.

      With these agents, AI models are now capable of not just responding to text input but also managing entire tasks. Today’s AI can plan, reason, and execute workflows across various tools and channels.

      OpenAI remains a leader in this space, closely partnered with Microsoft — yet competition has intensified. Google has advanced its Gemini models, Meta has developed LLaMA, and Claude comes from Anthropic. However, a trend emerges: the current leaders in the AI revolution are predominantly American tech giants who not only create and control the models but also the infrastructure and workflows that power businesses globally.

      The most significant competition arises from the East. China has illustrated through DeepSeek that it can catch up rapidly. This system has outperformed several Western models in benchmarks while operating at drastically lower computing costs. The substantial state support in Beijing ensures that China will not remain a follower for long.

      State of Europe

      Long before ChatGPT became widely recognized, European researchers established the foundations for modern AI.

      In the late 1980s, Yann LeCun (Université Pierre-et-Marie-Curie, Paris) was a pioneer in developing Convolutional Neural Networks (CNNs), which are essential for computer vision and multimodal AI. In 1997, Sepp Hochreiter and Jürgen Schmidhuber (TU Munich) created Long Short-Term Memory (LSTM) networks, pivotal for speech recognition, translation, and natural language processing.

      In 2016, London-based DeepMind shocked the world when its AlphaGo defeated world champion Lee Sedol; concurrently, LMU Munich’s CompVis group, led by Professor Björn Ommer, advanced generative AI, leading to Stable Diffusion.

      The paradox is evident: Europe innovates, while others commercialize. LeCun transitioned to Meta, DeepMind was acquired by Google, and Stable Diffusion was monetized through Stability AI in London and the US.

      A new generation of European AI companies aims to change this narrative. France’s Mistral is asserting Europe’s position in sovereign AI, developing open-source LLMs and enterprise tools like Le Chat that have attracted clients such as AXA and BNP Paribas. Black Forest Labs, formed from the same LMU team behind Stable Diffusion, is creating one of the most powerful image generators, intent on keeping cutting-edge research and commercialization within Europe.

      Other startups, such as Germany’s Langdock, have adopted a different strategy. Rather than building foundational models themselves, they facilitate businesses in utilizing existing LLMs while maintaining their data and workflows in-house.

      The forthcoming rise of AI employees

      We are merely beginning to uncover what AI is capable of. Today’s models — as robust as they are — are still in their infancy. The true value will not stem from the models individually, but from their application: automating workflows, enhancing operations, and effectively creating AI employees that can manage intricate tasks across various industries.

      For Europe, this presents a tremendous opportunity. Our industries excel in efficiency, supported by clearly defined, rule-based workflows — precisely the environment in which AI can excel. Manufacturing, logistics, finance, insurance, and even customer support possess substantial potential for enhancement or replacement by AI systems capable of adhering to documented procedures, making informed decisions, and carrying out entire processes independently.

      The monetization potential is vast. Every workflow eligible for automation represents an opportunity to decrease costs, boost speed, and scale operations. And unlike the consumer-driven hype surrounding AI, this value is concrete: it resides within enterprise productivity and efficiency.

      However, time is of the essence. If Europe does not act swiftly to train, deploy, and integrate AI employees locally, we risk becoming mere consumers. AI labor will be imported from the US, along with the corresponding value creation. Instead of spearheading the next industrial revolution, we will find ourselves compensating