The commercial rollout of AI applications is fundamentally transforming online retail. AI chatbots such as ChatGPT, Gemini, or Perplexity already advise consumers on their everyday purchases, compare products, and make recommendations. The first AI agents go a step further, autonomously handling search, comparison, and (US market only) purchase processes. The shift, however, also brings new challenges and risks for consumers: recommendations are personalised, delivered opaquely, and often not traceable for users.

Commissioned by the Austrian Federal Ministry of Labour, Social Affairs, Health, Care and Consumer Protection, the study examines how assistive AI (LLM-based chatbots) and executing AI (agentic systems) are transforming search, selection, and purchase processes in online retail, what risks this poses for consumers, and which regulatory questions are gaining relevance. Drawing on a literature review, illustrative practical tests, and a legal assessment, the study delivers a tangible overview of current technological developments while pointing to key fields of action for consumer protection.