Issue Brief

Institute for Future Strategy Issue Brief Vol. 63
“Transparency in AI Recommendations: Is ‘More Explanation’ the Answer? Policy Recommendations for Transparent AI Recommendations”

This issue brief examines, through empirical research, how to achieve transparency in AI recommendations—an increasingly important concern amid the spread of generative AI and personalized services—and proposes policy directions.

As platform recommendation algorithms exert a growing influence on consumers’ information searches and purchasing decisions, explaining why recommendations are made and what data they rely on has become a key policy challenge. This brief highlights that providing more explanation does not necessarily lead to more positive consumer responses.

The empirical findings show that making the reasons behind recommendations more specific does not consistently improve purchase intentions or attitudes toward platforms. By contrast, the types of data used to make recommendations have a clear effect on consumer responses. The same explanation can also be received differently depending on users’ characteristics and the context of use, including their shopping goals, privacy concerns, and AI literacy. This underscores the need for explanations tailored to the context rather than uniform standards.

Accordingly, this issue brief recommends shifting AI recommendation transparency policy from providing “more” explanation to providing “more appropriate” explanation. Platform services should clearly inform users that recommendations are being made and what data they rely on, offer explanations suited to the service context and users’ characteristics, and provide meaningful options to review, correct, or reject information. By tailoring these measures to the service and its users, platforms should implement AI transparency in a way that strengthens users’ actual understanding and freedom of choice, going beyond simple “information disclosure.”