Countries and economic blocs across the globe are at various stages of regulating artificial intelligence, ranging from a “Wild West” approach in the United States to highly intricate regulations in the European Union.
Here are some key points about AI regulation in major regions, leading up to the Paris AI summit on February 10-11:
United States
In December 2023, President Donald Trump reversed President Joe Biden’s executive order on AI oversight, issued in October 2023. The order, which was largely voluntary, required major AI developers like OpenAI to provide safety assessments and essential information to the federal government.
Supported by key tech companies, it aimed to protect privacy, prevent civil rights violations, and establish safeguards for national security. Currently, the United States has no formal AI regulations, although some existing privacy laws still apply.
As Yael Cohen-Hadria, a digital lawyer at consultancy EY, notes, under Trump, “it’s a complete Wild West,” with the administration effectively abandoning any formal regulation, allowing AI systems to operate freely.
China
China is still working on a formal law for generative AI but has introduced “Interim Measures” that require AI to respect personal and business interests, avoid using personal data without consent, flag AI-generated content, and protect users’ health.
Additionally, AI must “adhere to core socialist values,” which includes prohibitions on criticizing the Communist Party or undermining national security. While China closely monitors businesses, especially foreign ones, Cohen-Hadria predicts the government will grant itself significant exemptions from its own rules.
European Union
In stark contrast to the United States and China, the EU places a strong emphasis on protecting citizens’ rights in its AI regulation. The “AI Act,” which passed in March 2024, is the most comprehensive AI regulation globally, with certain provisions taking effect this week.
The law bans AI systems that engage in predictive policing based on profiling or use biometric data to infer personal traits like race, religion, or sexual orientation. The EU’s approach is risk-based: companies deploying high-risk AI systems face stricter obligations.
EU leaders argue that clear, comprehensive regulations will benefit businesses by improving intellectual property protections and facilitating the free flow of data.
India
India, like China, has a law on personal data but lacks specific legislation governing AI. The government has addressed harms from generative AI using existing laws related to defamation, privacy, copyright, and cybercrime.
While India recognizes the importance of its tech sector, it has yet to implement concrete AI regulations. Earlier in 2024, the government issued an advisory stating that firms would need government permission to deploy “unreliable” or “under-tested” AI models.
This followed a controversy involving Google’s Gemini, which accused Prime Minister Narendra Modi of implementing fascist policies.
United Kingdom
Britain’s Labour government has made AI a priority in its plan to stimulate economic growth. The UK is home to the world’s third-largest AI sector, after the United States and China. In January 2024, Prime Minister Keir Starmer unveiled an “AI opportunities action plan,” aiming for the UK to carve out its own regulatory path.
The plan emphasizes testing AI before regulating it, with the goal of ensuring that well-designed regulation fosters rapid and safe AI development. The UK is also conducting a consultation to clarify how copyright law applies to AI, aiming to protect the creative sector.
International Efforts
The Global Partnership on Artificial Intelligence (GPAI), which includes over 40 countries, seeks to promote responsible AI use.
Its members will meet on Sunday to discuss an “action plan for 2025.” Additionally, in May 2024, the Council of Europe adopted the first-ever binding international treaty on AI, with signatories including the US, UK, and the EU.
Of the 193 UN member countries, only seven are part of major AI governance initiatives, while 119, mostly in the Global South, belong to none.
