IP Law Daily, TECHNOLOGY/INTERNET NEWS: European Union’s AI Act enters into force, takes risk-based regulatory approach, (Aug 1, 2024)
Member states will have a year to designate oversight authorities to enforce the new law, which the EU calls the first legal framework regulating artificial intelligence technologies.
The European Union’s AI Act—billed as the world’s first-ever legal framework on artificial intelligence technology entered into force on August 1. The new law establishes harmonized rules on the regulation of AI in the EU. According to the European Commission, the AI Act “provides AI developers and deployers with clear requirements and obligations regarding specific uses of AI while reducing administrative and financial burdens for businesses.” Among other provisions, the Act bans certain AI applications based on sensitive characteristics, such as biometric categorization systems, and untargeted scraping of facial images from the internet or CCTV to create a facial image database. It also imposes clear obligations on other “high-risk” AI systems, including its use in critical infrastructure, education, employment, health care, banking, and justice and democratic processes like elections. In addition, it imposes transparency requirements on general purpose AI systems. A media release explains that EU member states have until August 2, 2025, to designate national authorities to oversee the application of the rules for AI systems. Text of the law is available here.
“The adoption of the AI Act is a significant milestone for the European Union,” said Mathieu Michel, Belgium’s secretary of state–digitization, in a news release issued when the Act was approved in May. “This landmark law, the first of its kind in the world, addresses a global technological challenge that also creates opportunities for our societies and economies. With the AI Act, Europe emphasizes the importance of trust, transparency and accountability when dealing with new technologies while at the same time ensuring this fast-changing technology can flourish and boost European innovation.”
Risk-based approach. The Act takes a tiered approach based on varying levels of risk. AI systems presented little or no risk of harm resulting from lack of transparency can be used freely. Examples of such applications are AI-enabled video games or spam filters.
Limited risk applications. AI systems presenting only a limited risk associated with lack of transparency in AI usage would be subject to light transparency obligations. For example, users of AI chatbots would have to be made aware that they are interacting with a machine, and AI-generated content must be identified as such.
High-risk applications. High-risk AI systems would include technology used in critical infrastructures, educational or vocational training, hiring and management of workers, essential public services like credit scoring, and certain law enforcement functions. These uses are authorized under the Act but “are subject to strict obligations before they can be put on the market.” These obligations include adequate risk assessment and mitigation systems; high quality of training datasets; logging of activity to ensure traceability of results; detailed documentation for compliance assessment; clear and adequate information to the deployer of the technology; appropriate human oversight measures; and a high level of robustness, security, and accuracy. The Commission said that such uses as cognitive behavioral manipulation and social scoring will be banned due to the unacceptable risk they pose. The law also prohibits the use of AI for predictive policing based on profiling and systems that use biometric data to categorize people according to specific categories such as race, religion, or sexual orientation.
Consequences for failure to comply. Companies not complying with the rules will be fined, the Commission said. Fines could be as high as seven percent of the global annual turnover for violations of banned AI applications, up to three percent for violations of other obligations, and up to 1.5 percent for supplying incorrect information to regulators.
AI Office to support implementation. On May 29, the European Commission announced its establishment an AI Office with the key goal of working “to foster research and innovation in trustworthy AI and position the EU as a leader in international discussions.” The Office will support the development and implementation of international agreements on AI, the Commission said. The Office will support Member States’ governance bodies and will directly enforce the rules for general-purpose AI models. The Commission said that the AI Office will coordinate the drafting of codes of practice, conduct testing and evaluation of general-purpose AI models, request information, and apply any necessary sanctions.
News: AINews TechnologyInternet GCNNews