Examining Transparency Requirements for AI-Based Chatbots and Websites

Abstract

<p>In this study, we analyze transparency of chatbots on Finnish well-being services counties’ websites. We study how well the transparency obligations under the EU Artificial Intelligence Act (AI Act) becoming applicable in August 2026, and the General Data Protection Regulation are met. We examined 12 chatbots and websites to assess whether they appropriately inform the user of the AI use and follow the transparency requirements of GDPR. The results indicate that, at this point in time, users are not informed transparently about the AI use. Websites also were not transparent enough about the personal data they send to third parties. These shortcomings highlight the need for clearer instructions and better design practices, so that transparency is achieved in the web-based social and health care services and the chatbots they use.<br></p>