Legal & Regulatory Guardrails for “Lawful Enterprise AI”
摘要
As GenAI becomes more prominent in business and society, global policymakers are racing to legislate around its possible negative impacts. This chapter looks at legal and regulatory issues about corporate compliance in four fundamental pillars: AI taxonomy, intellectual property, data management, and liability of products and services. It describes the landmark EU AI Act as the world’s first comprehensive, risk-based regulatory framework, contrasting it with the sectoral, pond-hopping approach of the USA. It covers the explosion of US state-level AI bills and legislative activity in other jurisdictions and aspires to international responsiveness. A deep dive into IP issues reveals a lack of consensus surrounding authorship and fair use of training data, while data governance deals with privacy, security, localization, and cross-border transfer obligations. It also looks at two principal AI product liability paradigms: those based on fault and those based on strict liability. With the maturing of legal frameworks, enterprises must remain proactive, ensuring compliance is built into the AI development lifecycle so that innovation remains within global legal frameworks.