世界互联网大会20余场分论坛齐举办，各界共议AI发展痛点

Core Insights - The 2025 World Internet Conference in Wuzhen focused on global development initiatives, digital economy, data governance, and AI's role in scientific research, featuring over 20 sub-forums with industry leaders discussing cutting-edge technology trends in the internet sector [1] Group 1: AI in Healthcare - The newly established "Smart Health Forum" at the conference addressed global health challenges, with experts emphasizing AI's potential to solve critical health issues [1] - Ant Group's CEO highlighted the AI health application AQ, which utilizes a medical large model to enhance healthcare services, including proactive questioning and health record management, with plans for further expansion in AI medical applications [1] - A collaboration with Shanghai Renji Hospital has led to the development of an intelligent urology system, improving diagnostic accuracy for primary care physicians by 4% to 8% [1] Group 2: Risks and Ethical Considerations - The application of AI in healthcare must be approached with caution due to privacy concerns and the potential for errors in medical advice and decision-making [2] - The need for strict data ethics in medical AI was emphasized, as patient data involves sensitive information [2] - AI's inability to fully replace doctors was noted, with the current role being that of an assistant to enhance medical practices [2] Group 3: AI Integration in Business - Internet companies must embed AI deeply into their business models and workflows to achieve sustainable growth and competitive advantage, as mere technological leadership is no longer sufficient [2] - AI was a recurring theme across various forums, highlighting its rapid development and the increasing focus on safety and governance issues [2] Group 4: AI Safety Governance - A proposed "AI Safety Governance Framework" was introduced, focusing on six dimensions: model safety governance, data and privacy, application and operation, foundational security, personnel and responsibility, and continuous improvement [3] - Emphasis was placed on managing four key aspects: framework, data, personnel, and supply chain to ensure AI safety throughout its lifecycle [3]