European Union Finalizes Historic AI Act to Regulate Artificial Intelligence
The European Union (EU) has reached a significant milestone in the regulation of artificial intelligence (AI) with the finalization of the AI Act, a groundbreaking piece of legislation. The AI Act establishes a comprehensive regulatory framework to address the risks associated with the rapid advancement of AI technology.
Classifying AI Uses by Risk Levels
The framework, proposed in 2021, categorizes AI applications based on their risk levels. Higher-risk applications will face stricter regulations to ensure safety and protect individuals’ rights. The president of the European Parliament, Roberta Metsola, describes the law as a “balanced and human-centered approach” that is set to become the global standard.
Banning Harmful AI Practices
The AI Act prohibits harmful AI practices that pose a clear threat to people’s safety, livelihoods, and rights. It specifically bans the exploitation of vulnerable groups, the use of biometrics for law enforcement purposes, and the deployment of manipulative “subliminal techniques.” Facial recognition by law enforcement and governments will also face stringent restrictions, with potential fines of up to 7 percent of global sales for violating companies.
Transparency Obligations for AI Systems
The legislation introduces a risk-based approach, imposing light transparency obligations on AI systems presenting limited risk. This includes chatbots like OpenAI’s ChatGPT and technology that generates images, audio, or video content. These systems will be subject to new transparency requirements under the law.
Historic Milestone for AI Regulation
European Commissioner Thierry Breton hails the AI Act as a historic achievement, emphasizing its significance as the first set of clear rules for AI use on a continent. The rules will not take effect until 2025, allowing room for technological advancements.
Global Concerns and Actions
The regulation of AI is a global concern. In late October, President Joe Biden signed an executive order to support responsible AI development while protecting the public from its potential dangers. The executive order introduces new standards for safety, privacy protection, equity, and civil rights in AI applications.
Industry leaders and experts have also expressed concerns about the risks posed by AI. More than 350 executives, researchers, and policymakers signed an open letter warning about the potential existential threat of AI developments. Lower-risk concerns have been raised by actors, writers, artists, and musicians, particularly in Hollywood, where regulations on AI use have been introduced.
The European Union’s AI Act sets a precedent for global AI regulation and aims to strike a balance between innovation and safeguarding individuals’ well-being. With its comprehensive framework, the EU is leading the way in addressing the challenges and controversies surrounding AI adoption.
