Regulatory Alerts2026-07-01
Added · Updated
Köhler-Geib Interview: Europe Must Strengthen Digital Sovereignty
Bundesbank board member Fritzi Köhler-Geib warns that Europe risks falling behind the US and China in artificial intelligence, necessitating a push for digital sovereignty to ensure access to critical models. She highlights the dangers of algorithmic herding behavior in financial markets caused by AI training on identical datasets and notes that while central banks are actively monitoring these risks, AI agents are increasingly making autonomous trading decisions. Köhler-Geib further cautions that current AI models lack self-reflection and cannot reliably eliminate biases, representing a significant limitation for their use in finance.
© Gaby Gerster
01.07.2026
Artificial intelligence is a disruptive technology that is profoundly changing the economy and society. In Europe, we must make greater use of the opportunities that arise from this, says Bundesbank board member Fritzi Köhler-Geib in an interview with the Süddeutsche Zeitung. In the interview, she discusses the opportunities and challenges that AI brings for Europe and the stability of financial markets.
Europe Lagging Behind in AI
According to Köhler-Geib, Europe is lagging behind the US and China in large AI models. As a current example, she cites the US company Anthropic, whose latest AI models were blocked for the rest of the world by order of the US government.
This clearly shows the strategic relevance these technologies have acquired, emphasizes the board member. If Europe does not have access to models with such capabilities, it must develop its own alternatives:
We must advance our digital sovereignty.
AI in Financial Markets
The use of AI in financial markets carries both opportunities and risks. What is crucial is which models are used and with which data they are trained.
If all models are based on the same datasets, this quickly leads to a kind of convergence, explains Köhler-Geib. This could reinforce herding behavior:
Transferred to humans, this would be like all stock investors having the same education, the same models, and making identical decisions.
The phenomenon of herding behavior is well known to markets. However, through AI, it is gaining a new technological dimension.
To counter such risks, central banks and supervisory authorities are already active.
They are trying to get an early picture, understand possible risks, and – where necessary – set guardrails, says Köhler-Geib. An example of this is the Logos project of the BIS Innovation Hub, in which central banks analyze the behavior of AI agents in a simulated environment.
New Generation of Trading Algorithms
Algorithmic trading has been established for decades, but AI is shifting the boundary up to which humans have direct influence on trading decisions.
With so-called 'agentic' AI, which is also becoming increasingly powerful, systems are making more and more independent decisions without human intervention in individual cases, explains Köhler-Geib in the interview.
The European Banking Authority is currently working to gain a better overview of where and how AI is used in trading. The risks of these models do not necessarily have to be greater than those from classical algorithmic trading, says the Bundesbank board member.
Limits of AI Models
According to Köhler-Geib, biases can creep into AI models, similar to human investors:
These models can often identify such biases excellently. However, they cannot yet reliably eliminate them.
This is also due to the fact that these models lack consciousness or self-reflection.
This can continue to develop in the coming years – but for now, this remains a central limitation in dealing with AI in financial markets, she says.