AI in finance in Germany

AI in German finance covers a wide range of tasks, from a chatbot that answers a routine account question to a model that flags a suspicious card payment within seconds. In EU securities markets, most reported uses of AI help staff draft and summarize text, with a person checking the output before it reaches a client, and asset managers such as those based in Munich or Frankfurt work the same way. On the insurance side, firms including Allianz and Munich Re, both headquartered in Munich, run data science teams that build pricing and claims models, work that sits under the same EU rules as banking AI once it reaches the high-risk categories in insurance.

Fraud detection is one of the older, more established uses: a model that scores card payments in real time and blocks the unusual ones is common across German banks, and it sits outside the high-risk categories of the EU AI Act, which names credit scoring and insurance pricing instead.

BaFin published its principles for the use of algorithms in decision-making processes before the EU AI Act existed, and they still apply on top of it. BaFin splits a model's life into two phases: development, where it is selected, calibrated and validated, and application, where staff interpret and act on its output, and it asks every institution to name a clear owner for each phase. Under the German Act Implementing the European Artificial Intelligence Act, BaFin now also supervises AI that banks and insurers use in direct connection with regulated activities, such as customer chatbots and creditworthiness checks, while AI outside those activities, such as HR software, falls to the Bundesnetzagentur.

Two European authorities in Frankfurt add sector-specific guidance. ESMA told investment firms that MiFID II duties apply in full when they use AI, and named algorithmic bias and overreliance on AI as risks to watch. EIOPA published an Opinion on AI governance and risk management covering the AI insurers use in pricing, underwriting and claims.

Finance Loop is the meeting place for people who build, buy and supervise AI at banks, insurers and fintechs across Germany. It connects the finance, IT and AI communities in Frankfurt, Munich, Berlin and Hamburg.

Finance Loop announced KI Exchange 2026 in Hamburg, a conference by Payment & Banking whose program covered sovereign AI infrastructure, AI governance and DORA compliance, with speakers expected from BaFin. It highlighted fAInance near Frankfurt Airport, a conference by Sopra Steria and Fraunhofer IAIS for staff of financial institutions. In Frankfurt, Finance Loop presented an official side event of AI Week Frankfurt 2025 with Frankfurt Main Finance and AI Hub Frankfurt, with talks on AI against financial crime and on AI testing. Its cooperation with Frankfurt Data Science covers AI governance and applied data science in financial institutions, and it supports Claude Hacker House, a hands-on series on Claude for business and finance use cases in Frankfurt, Berlin and Munich. Frankfurt Digital Finance had GenAI on the program of its first day, and AI and enterprise adoption is one of four topic areas of CONF3RENCE in Dortmund. Finance Loop also has a Munich chapter that holds its own meetups on digital assets and finance technology. Related pages: AI in finance in Frankfurt, AI in banking in Germany and the EU AI Act in finance.

Most AI in German banking supports staff work: drafting text, summarizing documents and answering routine questions, always with a person checking the result. Fraud detection on card and account payments is a longer-established use. Consumer credit scoring is the use that carries the most regulatory weight, since it counts as high-risk under the EU AI Act.

BaFin expects a named owner for a model at every stage of its life, from selection and calibration to the point where staff act on its output, following its own algorithm principles. Under the German Act Implementing the EU AI Act, BaFin also supervises AI that supervised firms use for regulated activities such as credit scoring and customer chatbots.