HEC Paris and Goethe University Frankfurt researchers find: Adding a human touch makes the AI financial advice more persuasive

As artificial intelligence continues to reshape the financial services industry, a new study suggests that combining human expertise with AI-driven services may significantly boost financial investment outcomes due to greater decision comfort provided by the presence of a human in the loop.

In both field and online experiments with one of Germany’s largest savings banks, academics from HEC Paris (Professors Cathy Liu Yang and Xitong Li) and Goethe University Frankfurt (Professors Kevin Bauer and Oliver Hinz) examined how human involvement in AI-based service affects both the upstream AI-based service and downstream consumers in the context of investment decisions. They found that, compared to pure AI financial advice, bankers acting as gatekeepers do not necessarily compromise the quality of AI-based financial advice, while increasing customers’ alignment with the advice.

The study, titled “My Advisor, Her AI and Me: Evidence from a Field Experiment on Human-AI Collaboration and Investment Decisions”, examined real-world investment decisions among 137 bank customers on personal loans. The results were clear: clients who received human-AI collaborative advice aligned their decisions more closely with the recommendations, especially in high-risk scenarios, and achieved better financial returns.

The study took a two-pronged approach, combining a field experiment with a controlled online experiment to investigate both the production and consumption aspects of human-AI collaborative investment advice.

Key Findings

The researchers discovered that customers were much more likely to follow investment advice when it came from a human-AI collaboration rather than AI alone, especially in riskier decisions—leading to a 15.5 percentage point rise in alignment and better financial returns. This effect was not because of perceived improvements in advice quality but because of the emotional reassurance provided by human involvement.

“Contrary to concerns that human intervention might reduce AI’s effectiveness, our research shows that involving a human banker not only preserves the quality of advice, but actually enhances its persuasive power, primarily driven by emotional trust,” said lead author Professor Cathy Yang of HEC Paris.

Further analysis showed that the presence of a human increasing emotional trust and thus persuasiveness under decision uncertainty—not because of perceived improvements in the advice quality due to complementarity of human-AI collaboration. “In uncertain situations, people value decision comfort and emotional assurance—something AI alone cannot offer,” said co-author Professor Xitong Li of HEC Paris.

The findings come at a critical time, as regulators, including the EU, are discussing how much human oversight should be required in AI-driven services. The study advocates a consumer-focused approach, recommending policies that promote effective human-AI collaboration to help protect consumer welfare while improving outcomes. 

“This work provides vital insights for banks, tech providers, and regulators alike,” concluded Professor Yang. “It shows that ‘human in the loop’ isn’t just about compliance or control—it can be a strategic advantage.”

Yang, Cathy L. and Bauer, Kevin and Li, Xitong and Hinz, Oliver, My Advisor, Her AI and Me: Evidence from a Field Experiment on Human-AI Collaboration and Investment Decisions (June 04, 2025). Available at SSRN: https://ssrn.com/abstract=5281742 or http://dx.doi.org/10.2139/ssrn.5281742