European Union officials have reached an agreement on the Artificial Intelligence Act, a set of regulations designed to govern the use of artificial intelligence (AI) in the 27-nation bloc. The Act takes a "risk-based approach" and focuses on regulating uses of AI rather than the technology itself. Its aim is to protect democracy, the rule of law, and fundamental rights while encouraging investment and innovation. The legislation categorizes AI applications based on their level of risk, with high-risk systems facing stricter requirements, such as using high-quality data and providing clear information to users. Certain AI uses, such as social scoring systems and emotion recognition systems in schools and workplaces, are banned due to their deemed unacceptable risk.

The AI Act will not take effect until two years after final approval from European lawmakers, which is expected in early 2024. Violations of the Act could result in fines of up to 35 million euros ($38 million) or 7% of a company's global revenue.

Although the Act will apply to the EU's nearly 450 million residents, its impact is expected to extend beyond the bloc, with the EU's regulations potentially serving as a global standard. The EU has previously played a leading role in setting tech directives that have had global repercussions, such as the mandating of a common charging plug that forced Apple to abandon its Lightning cable. Experts believe that the AI Act will significantly contribute to the global momentum to regulate AI across jurisdictions and place the EU in a unique position to lead the way in governing AI.

However, rights groups have criticized the AI Act for not pursuing a full ban on live facial recognition, which they argue enables dystopian digital surveillance. Amnesty International expressed concerns about the precedent set by the partial ban and called for a ban on the export of AI technologies that can harm human rights, such as social scoring.

The United States and China, the world's major AI powers, have also initiated their own regulations on AI. President Joe Biden signed an executive order on AI in October, requiring leading AI developers to share safety test results with the government. China has released "interim measures" for managing generative AI and proposed a Global AI Governance Initiative.

The rise of OpenAI's ChatGPT technology prompted European policymakers to update their proposal. The AI Act includes provisions for chatbots and general-purpose AI systems, imposing transparency requirements and compliance with EU copyright law. Stricter rules apply to the most advanced AI systems to prevent systemic risks.

Overall, the AI Act represents a significant step in regulating AI in Europe and has the potential to shape global AI governance. Its impact will be felt beyond the EU, with other countries closely observing its provisions and potentially adopting similar regulations.