Governments across the globe are grappling with how to regulate artificial intelligence, and the world's three leading AI powers—the European Union, the United States, and China—are pursuing sharply different strategies. While the concerns are similar—AI-enabled cyberattacks, national security threats, and the impact of chatbots on vulnerable users—the policy responses could hardly be more divergent.
EU: Binding rules with a pragmatic twist
The European Union adopted its AI Act in 2024, a landmark regulation that Brussels describes as the world's first comprehensive framework for AI. The law classifies AI systems by risk, from minimal-risk applications like spam filters to unacceptable uses such as social scoring and emotion recognition, which are outright banned. High-risk systems face strict requirements, though the EU has already shown flexibility: under pressure from industry, the AI Omnibus bill delayed some obligations for high-risk systems, reflecting a shift to the right in the European Commission after the EU elections.
European Commission President Ursula von der Leyen has defended the approach, saying the rules put "Europe in the position to shape global efforts." She has also invited leading AI labs to discuss how to support industry efforts to pace the frontier. However, European companies worry that over-regulation could hamper their ability to compete with US and Chinese rivals, and some suspect that calls for caution are a ploy by American tech giants to cement their dominance.
The EU is also using other tools. In August, ChatGPT became the first AI chatbot added to the EU's list of very large digital services under the Digital Services Act, subjecting it to greater legal scrutiny. In September, Brussels proposed the EU Kids Act, which would require companies to make AI companions and chatbots safe for children before they are deployed. The proposal specifically targets chatbots that "simulate interpersonal relationships in ways that create emotional dependency," following a series of suicides linked to such interactions.
US: Voluntary pledges and an iron grip on language
The United States, under President Donald Trump, has taken a markedly different path. On his first day back in office, Trump rescinded Joe Biden's 2023 executive order that required developers of powerful AI systems to share safety test results with the government. Instead, the White House unveiled an "AI Action Plan" in July 2025, listing more than 90 measures to accelerate development, boost infrastructure investment, and deregulate the sector.
An accompanying executive order, titled "Preventing Woke AI in the Federal Government," mandates that federal agencies purchase only AI models deemed ideologically neutral. The order lacks a formal methodology for assessing neutrality, but companies like Anthropic and xAI have been eager to comply, sharing technical documentation with the government.
Congress, deeply divided and heavily lobbied by Big Tech, has yet to pass a federal AI law. State laws, such as California's AI Transparency Act, are taking effect despite Trump's attempts to block them. In June, a new executive order established a voluntary system for developers to give the government up to 30 days of early access to their most powerful models, a response to growing concerns about AI-enabled cyberattacks.
On September 29, the heads of America's six largest AI companies signed a White House accord pledging independent audits and board-level oversight of cutting-edge models. The pact is self-regulatory and carries no penalties for non-compliance, though Trump called it "morally binding." The same day, he ordered federal agencies to replace the term "artificial intelligence" with "super intelligence" in official communications, a symbolic move that underscores his administration's focus on maintaining US dominance.
The economic stakes are high: AI infrastructure and investment currently account for roughly a fifth to a third of US economic growth, contributing between 0.45 and 0.97 percentage points to overall growth, according to the Peterson Institute for International Economics.
China: State control and human oversight
China has built a complex and evolving regulatory framework for AI, with a strong emphasis on state control. Regulations on AI companion chatbots took effect in July, aiming to curb emotional dependency and banning the creation of digital clones of real people without consent. President Xi Jinping has stated that AI should be "always under human control," calling for more laws, technological monitoring, early warning systems, and emergency response mechanisms.
Premier Li Qiang has also warned that the world risks "losing control" of frontier technologies if governments are too slow to regulate them. However, China has rejected calls for a general pause or deliberate slowdown in AI development, preferring to advance its own capabilities while maintaining tight oversight.
The three approaches reflect different priorities: the EU seeks to balance innovation with fundamental rights, the US prioritizes economic growth and technological leadership, and China emphasizes state security and social stability. As AI continues to evolve, the global regulatory landscape will likely remain fragmented, with each bloc charting its own course.
