<p>The transition from traditional keyword-based search engines to generative AI-powered search systems has introduced significant legal and governance challenges. AI-driven search models are reshaping how knowledge is produced, shared, and evaluated, raising fundamental questions around diversity, cognitive trust, autonomy, responsibility, and social rationality. Previously, search services focused on issues like theft and personalized ranking, but generative search blurs the lines between content searchers and creators. While existing research primarily addresses copyright concerns related to training data or personalized ranking systems, it does not fully explore the responsibility for reusing copyrighted content in generated results, or issues like filter bubbles and fake news creation. To address these challenges, this study combines theoretical and policy analysis, referencing recent cases like New York Times v. OpenAI and Walters v. OpenAI, and the latest policies such as the EU’s AIA Act 2024 and China’s “raw forming artificial intelligence” bill. It delves into the implications of piggybacking responsibility, personality control, and the liability of virtual products. The findings reveal that AI search is a blend of query and innovation, increasing the risk of intellectual property infringement, personalization, and new liability claims. This research offers a comprehensive management system integrating intellectual property rights, risk management, and algorithmic accountability, providing vital guidelines for policymakers, legal professionals, and the public to safeguard trust in digital information systems.</p>