EU agrees rules for artificial intelligence with groundbreaking new law
European Union policymakers on Friday agreed to a sweeping new law to regulate artificial intelligence, one of the world’s first comprehensive attempts to limit the use of a rapidly evolving technology that has far-reaching social and economic implications.
The law, called the AI law, sets a new global benchmark for countries looking to harness the technology’s potential benefits while protecting themselves from its potential risks, such as automating jobs, spreading misinformation online and endangering national security. The law still needs to go through a few final steps before it is adopted, but thanks to the political agreement, its main contours have been laid down.
European policymakers focused on the riskiest uses of AI by companies and governments, including those for law enforcement and the operation of crucial services such as water and energy. Makers of the largest general-purpose AI systems, such as those powering the ChatGPT chatbot, would face new transparency requirements. Chatbots and software that creates manipulated images such as ‘deepfakes’ should make it clear that what people saw was generated by AI, according to EU officials and previous draft laws.
The use of facial recognition software by police and governments would be restricted, subject to certain exceptions in the areas of security and national security. Companies that violate regulations can face fines of up to 7 percent of global turnover.
“Europe has positioned itself as a pioneer and understands the importance of its role as a global standard setter,” said Thierry Breton, the European Commissioner who helped lead the negotiations the agreementsaid a statement.
But even as the law was hailed as a regulatory breakthrough, questions remained about its effectiveness. Many aspects of the policy were not expected to come into effect for twelve to twenty-four months, which is a significant amount of time for AI development. And until the last minute of the negotiations, policymakers and countries fought over the language of policy and how to balance promoting innovation with the need to protect against potential harm.
The deal reached in Brussels took three days of negotiations, including an initial 22-hour session that started Wednesday afternoon and dragged into Thursday. The final agreement was not immediately made public as talks were expected to continue behind the scenes to finalize technical details, which could delay final passage. To vote must be kept in the Parliament and the European Council, which consists of representatives of the 27 countries of the union.
Regulating AI took on urgency after last year’s release ChatGPT, which became a global sensation by demonstrating the advancing capabilities of AI. In the United States recently the Biden administration issued an executive order partly focused on the effects of AI on national security. Britain, Japan and other countries have taken a more hands-off approach, while China has imposed some restrictions on data use and recommendation algorithms.
At stake trillions of dollars in assessed value as AI is predicted to reshape the global economy. “Technological dominance precedes economic dominance and political dominance,” says Jean-Noël Barrot, France’s Digital Minister. said this week.
Europe is one of the countries that has made the furthest progress in regulating AI, having started working on what would become the AI Act in 2018. In recent years, EU leaders have sought to bring a new level of oversight of technology, similar to healthcare regulation. or banking sectors. The region has already done that far-reaching laws were passed related to data privacy, competition and content moderation.
a first design of the AI Act was released in 2021. But policymakers had to rewrite the law as technological breakthroughs emerged. The first version made no mention of general-purpose AI models, such as those powering ChatGPT.
Policymakers agreed to what they called a “risk-based approach” to regulating AI, with a defined set of applications facing the most scrutiny and restrictions. Companies that create AI tools that have the most potential harm to individuals and society, such as in hiring and training, should provide regulators with evidence of risk assessments, breakdowns of what data was used to train the systems, and assurances that the software did. not cause harm, such as perpetuating racist prejudices. Creating and implementing the systems would also require human supervision.
Some practices, such as the random scraping of images Using the Internet to create a facial recognition database would be completely banned.
The debate in the European Union was contentious, a sign of how AI has confused lawmakers. EU officials were divided over how deeply to regulate the newer AI systems, fearing a handicap for European startups trying to overtake U.S. companies like Google and OpenAI.
The law added requirements for makers of the largest AI models to release information about how their systems work and to evaluate for “systemic risks,” Mr. Breton said.
The new regulations will be closely monitored worldwide. They will affect not only major AI developers such as Google, Meta, Microsoft and OpenAI, but also other companies expected to use the technology in areas such as education, healthcare and banking. Governments are also increasingly turning to AI in criminal justice and the allocation of public benefits.
Enforcement remains unclear. The AI Act involves regulators from 27 countries and calls for hiring new experts at a time when government budgets are tight. Legal challenges are likely as companies test the new rules in court. Previous EU legislation, including the landmark digital privacy law known as the General Data Protection Regulation, has been criticised unevenly implemented.
“The EU’s regulatory capacity is in question,” said Kris Shrishak, a senior fellow at the Irish Council for Civil Liberties, who has advised European lawmakers on the AI Act. “Without strong enforcement, this deal will have no meaning.”