AI in finance in Munich

Munich's finance sector grew out of insurance. Allianz and Munich Re are headquartered in the city, next to Generali Deutschland and the German operations of Swiss Re, and InsurTech Hub Munich, backed by Allianz, Generali and Munich Re, works with startups on claims automation and fraud detection. The fintech in Munich page tells that story in more detail.

So when people search for AI in finance in Munich, most of the work they find sits in claims, underwriting and pricing. Allianz runs its Insurance Copilot for motor claims in Austria and has extended it to property claims: the tool gathers data from the claim and the contract, reads documents and photos, flags gaps and drafts emails, and the claims handler decides on cover and payout. Banking and asset management are the smaller part of the local picture. Scalable Capital, the robo-advisor and broker founded in Munich, and the asset management arm of Munich Re both hire the same quantitative and data people, as the WealthTech in Munich page describes.

Munich Re also approaches AI from the other side: it insures it. Its product aiSure is a performance guarantee for AI solutions. An AI vendor that has aiSure behind it can promise its clients compensation when a model makes errors, and the policy pays when the model falls short of a performance level agreed when the cover starts. Before Munich Re writes such a policy, the AI solution goes through a technical due diligence, and the guarantee holds only for the use cases defined in that review.

For a bank or insurer that buys AI, this changes the conversation with the vendor. A credit model or a claims model with a measured error rate and an insurer standing behind that rate is easier to take through a model risk committee. It does not move the regulatory duty, though: the bank that uses the model stays responsible for it toward BaFin.

The Munich Center for Machine Learning (MCML) is a joint initiative of LMU Munich and the Technical University of Munich (TUM). Founded in 2018, it is one of Germany's six national AI competence centers and now has permanent funding from the federal government and the state of Bavaria. More than 80 research groups work there on the foundations of machine learning, on vision and language, and on machine learning for specific domains. For a quant or a risk modeler, the domain work and the language models are the most relevant parts; students of both universities who take a machine learning course in finance often come through this network.

The appliedAI Institute for Europe is the second Munich address. It is a non-profit subsidiary of the appliedAI Initiative, which UnternehmerTUM and IPAI set up as a joint venture, and it was founded in Munich in 2022. The institute runs trainings, courses and events and publishes AI tools and research on governance, which makes it a place where compliance and IT teams of Munich banks and insurers learn how others put models into production.

Munich has no financial regulator of its own. BaFin supervises Munich banks and insurers from Bonn and Frankfurt, and under the German act implementing the EU AI Act it also supervises the AI that these firms use in direct connection with regulated business. The EU AI Act lists two finance uses as high-risk in Annex III: credit scoring of natural persons, and risk assessment and pricing in life and health insurance. The second one matters more in Munich than anywhere else in Germany, because so much life and health insurance is written here.

Insurers also answer to Solvency II and to the guidance of EIOPA, whose Opinion on AI governance and risk management covers AI in pricing, underwriting, claims and fraud detection. A claims copilot that only drafts emails is not high-risk under the AI Act, but it still runs on an IT service that belongs in the insurer's DORA register. The EU AI Act in financial services page goes through the duties one by one.

Finance Loop is the meeting place for actuaries, data scientists, compliance people and founders who work on AI in Munich's insurers, banks and asset managers. It launched its Munich chapter with an event in the city, is a network partner of W3MUC, the Munich web3 hub, and connects the finance, IT and AI communities across Germany, Austria and Switzerland, with events in Frankfurt, Berlin and Hamburg too.

With Finteda and Frankfurt Data Science, Finance Loop supports Claude Hacker House, a hands-on series on building with Claude for business and finance. Its organizers describe it as a series for financial centers such as Frankfurt, Berlin, Munich and Zurich. The Bitcoin seminar in Munich is the Finance Loop format for investors in the city who want to understand bitcoin as an asset class.

aiSure covers the performance of an AI solution. When a model fails to meet the performance level agreed at the start, the policy pays, so an AI vendor can compensate its clients for losses caused by model errors. Munich Re first checks the solution in a technical due diligence and limits the cover to the use cases defined there.

AI in finance in Munich mostly comes up inside broader AI conferences, insurance events and meetups. Finance Loop lists the dates it runs or supports in Munich in the calendar above and on the events page. The closest dedicated conferences are fAInance near Frankfurt Airport, run by Sopra Steria and Fraunhofer IAIS for staff of financial institutions, and KI Exchange in Hamburg.

LMU Munich and the Technical University of Munich run the Munich Center for Machine Learning together, with more than 80 research groups. The appliedAI Institute for Europe, which goes back to UnternehmerTUM and IPAI, adds courses and trainings for people who already work in companies.

Finance Loop holds meetups in Munich with its chapter and with W3MUC, and Claude Hacker House lists Munich among its cities. The appliedAI Institute runs its own events and hackathons. For AI in finance dates across Germany, the Finance Loop events page is the place to check.