EU Artificial Intelligence Act
A regional artificial intelligence legislative framework adopted by the European Union in 2024. It is the world's first systematic AI legislation and adopts a risk-tiered approach to regulating AI applications.
Core Facts
The EU AI Act has extraterritorial effect: even if a company's headquarters is not within the EU, it must comply with the regulations as long as its AI system outputs are directed at the EU market or affect EU citizens.
Companies that violate the AI Act can face fines of up to 7% of their global annual turnover.
欧盟《AI法案》要求AI公司披露训练数据来源
Timeline (last 90 days)
AI security assessment currently lacks a globally unified mandatory standard, but framework documents such as the NIST AI Risk Management Framework, the EU AI Act compliance assessment requirements, and ISO/IEC 42001 have emerged.
The EU AI Act classifies deepfakes in the category subject to transparency obligations, requiring content that generates or manipulates images, audio, or video to clearly disclose that it is artificially synthesized, though scenarios such as artistic creation and satire may receive limited exemptions.
The transparency provisions of the EU AI Act mainly bind providers and deployers of systems offering services in the EU market, and have limited direct binding force on state actors outside the EU.
The EU's Artificial Intelligence Act faces two practical difficulties in regulating political advertising: limited extraterritorial effect and difficulty in technical determination.
AI regulatory approaches under discussion in the U.S. Congress primarily include banning or pausing the development of superintelligent AI, introducing external testers and oversight mechanisms, and holding AI companies legally and financially liable for harms.
The EU AI Act has extraterritorial effect: even if a company's headquarters is not within the EU, it must comply with the regulations as long as its AI system outputs are directed at the EU market or affect EU citizens.
All Facts (11)
The EU AI Act has extraterritorial effect: even if a company's headquarters is not within the EU, it must comply with the regulations as long as its AI system outputs are directed at the EU market or affect EU citizens.
80%VerifiedCompanies that violate the AI Act can face fines of up to 7% of their global annual turnover.
65%Verified欧盟《AI法案》要求AI公司披露训练数据来源
65%Unverified《人工智能法案》采用风险分级管理，将AI应用分为不可接受风险、高风险、有限风险和最小风险四个等级
75%Unverified欧盟《AI法案》已将深度伪造列为高风险应用，要求生成内容必须明确标注为AI生成
60%UnverifiedAI security assessment currently lacks a globally unified mandatory standard, but framework documents such as the NIST AI Risk Management Framework, the EU AI Act compliance assessment requirements, and ISO/IEC 42001 have emerged.
50%UnverifiedThe EU AI Act classifies deepfakes in the category subject to transparency obligations, requiring content that generates or manipulates images, audio, or video to clearly disclose that it is artificially synthesized, though scenarios such as artistic creation and satire may receive limited exemptions.
50%UnverifiedThe transparency provisions of the EU AI Act mainly bind providers and deployers of systems offering services in the EU market, and have limited direct binding force on state actors outside the EU.
50%UnverifiedThe EU's Artificial Intelligence Act faces two practical difficulties in regulating political advertising: limited extraterritorial effect and difficulty in technical determination.
50%UnverifiedAI regulatory approaches under discussion in the U.S. Congress primarily include banning or pausing the development of superintelligent AI, introducing external testers and oversight mechanisms, and holding AI companies legally and financially liable for harms.
50%Unverified欧盟《AI法案》采用风险分级立法路径，对通用目的AI模型设置透明度、版权合规与安全测试义务
50%