45.4% of all respondents trust their bank more than any other institution when it comes to handling AI data. Younger people trust their bank the most: 51.5% of 18-29 year olds believe their bank handles AI financial data securely.
- 33.4% of all respondents use popular AI tools such as ChatGPT or Gemini at least once a week
- 27.9% have their financial transactions explained to them by AI
- 81.2% are in favor of their AI financial data being stored only within the EU
Berlin – February 18, 2026 – AI is a new component of many processes in banking, such as fraud detection, the automation of internal processes, and chatbots. It is also known that consumers are increasingly using large language models such as ChatGPT to answer financial questions. However, the question remains: What expectations do customers have of AI in banking? How far does their trust extend when AI agents can access accounts, cards, or loans in the future? And what is the overall sentiment among consumers regarding AI in banking?
The results of the current survey show that AI is already part of everyday life for a significant portion of the German population: 33.4% of respondents use popular AI tools such as ChatGPT or Gemini at least once a week. There are hardly any differences between the sexes – 32.9% of women and 33.8% of men regularly use AI tools.
Many consumers already use AI specifically for financial matters. In the past twelve months, 27.9% had financial transactions explained to them by AI, and 27.8% used AI chatbots for banking matters. In addition, 19.8% acquired new financial knowledge with the help of AI, and 10.8% used AI tools to create their ETF portfolios. Younger user groups are particularly active: 41% of 18- to 29-year-olds use AI to explain their personal transactions. Among 30- to 39-year-olds, 15.6% use AI to create an ETF portfolio.
Despite its increasing use, security concerns remain widespread. 45.6% of respondents are concerned about potential data leaks, and 34.5% fear excessive data collection from AI tools. Younger respondents, who use AI most intensively, cite an above-average number of risks.
Trust is where opinions diverge: 45.4% of all respondents trust their bank most when it comes to handling AI-related financial data, followed by insurance companies with 40.6% – government institutions only achieve 37.6%. Among 18- to 29-year-olds, trust in their own bank is highest at 51.5%.
At the same time, respondents articulate strict requirements for data sovereignty – an indication of clear rules for the use of AI in banking and financial data. 84.7% want to be able to download or delete personal data at any time. 82.4% consider it important to be able to switch an AI agent on and off, and 81.2% are in favor of their data being stored only within the EU.
The results show that AI tools in the financial sector generally enjoy a high level of consumer trust. At the same time, financial institutions should develop these services responsibly themselves – with data stored in the EU and clear options for switching them on and off – instead of leaving the market entirely to providers from the US or China, whom many customers view with considerable skepticism.Steffen Jentsch, CEO Solaris SE
The results reveal a remarkable ambivalence: on the one hand, many consumers already actively use AI-supported functions for financial matters – in some cases even for sensitive personal data such as transaction analyses or portfolio recommendations. On the other hand, the majority express clear reservations about data security and transparency. It is also noteworthy that, despite high usage, younger age groups are the most likely to identify risks and have the least trust in government institutions. The data suggests that the acceptance of AI in banking depends largely on clear data protection mechanisms and traceable data flows.
Civey conducted an online survey for Solaris from November 14 to November 20, 2025. The survey covered 4,000 18- to 65-year-olds who use online banking in Germany. The results are representative of the target group surveyed, taking into account a statistical error of 2.5 percentage points in the respective overall result. Civey collects data using its own technology among verified participants.