Pressure is growing worldwide to rein in the most powerful artificial intelligence systems, and the world’s three top AI heavyweights are answering in three very different ways.
The worries are much the same everywhere, starting from AI-powered cyberattacks, threats to national security and chatbots linked to children’s suicides. The responses, it seems, are significantly different.
The EU has bet on binding rules, though it has already loosened some of them under pressure from industry. The US relies on voluntary promises and executive orders meant to keep its AI boom going. China is building its own rulebook under tight state control while rejecting calls to slow down.
Here is how each is going about it.
EU: “The world’s first comprehensive rules on AI”
The European Union adopted its AI Act in 2024, which Brussels calls the “world’s first comprehensive rules on artificial intelligence”.
The Act regulates AI based on a four-tier risk framework, with rules escalating according to the threat level the AI system poses, going from those posing limited to no threat — such as video game elements and spam filters being able to operate freely — all the way to unacceptable systems with wide-reaching powers being outright banned such as social scoring, emotion recognition and manipulation.
But European companies, fearing the rules would stifle innovation as they race to catch up with US and Chinese rivals, lobbied hard and later won a delay as part of the AI Omnibus bill on rules covering high-risk systems, thanks in part to the Commission’s shift to the right following the EU elections.
Amid growing global calls for control, European Commission President Ursula von der Leyen said last month that the AI rules put “Europe in the position to shape global efforts”.
She also said she would invite the main frontier labs “for a discussion on how we can support ongoing industry efforts to pace the frontier”.
European firms have been more wary of slowing down, fearing it could be an attempt by US giants to lock in their dominance.
The EU has other tools to regulate AI.
In August, ChatGPT became the first AI chatbot added to the EU’s list of very large digital services subject to greater legal scrutiny under its online content rules known as the Digital Services Act.
Brussels last month also proposed forcing companies to make AI companions and chatbots safe before children use them.
The EU Kids Act calls for banning AI chatbots from “simulating interpersonal relationships in ways that create emotional dependency,” the EU said, following a spate of suicides worldwide linked to children’s interactions with such assistants.
US: ‘Super intelligence’
Nearly two years into his second term, President Donald Trump has built his AI policy almost entirely from voluntary industry pledges and executive orders intended to sustain an AI boom that is powering the US economy.
On his first day back in office, Trump scrapped his predecessor Joe Biden’s 2023 executive order requiring developers of the most powerful AI systems to share safety test results with the government and directing federal agencies to set guidelines for privacy to prevent algorithmic discrimination.
In July 2025, the White House unveiled an “AI Action Plan” listing more than 90 measures to speed up development, ramp up investment in infrastructure and deregulate the US sector as a whole.
An accompanying executive order titled “Preventing Woke AI in the Federal Government” requires federal agencies to buy only models deemed ideologically neutral.
The order was issued without any formal methodology to assess their neutrality other than sharing their platform’s technical documentation with the US government, something Dario Amodei’s Anthropic and Elon Musk’s xAI were all too happy to do especially given the latter’s history of controversial incidents.
Congress, politically divided and heavily lobbied by Big Tech, has yet to pass a federal AI law and is out of session until after the midterm elections in November.
Trump has tried to block state AI laws, but Congress has repeatedly refused to override them, and state laws are taking effect nationwide including California’s AI Transparency Act.
This year, the release of AI models capable of powerful cyberattacks shifted the focus to security. A June executive order set up a voluntary system for developers to give the government up to 30 days of early access to their most powerful models.
With voter anxiety growing, the heads of America’s six largest AI companies signed a White House accord on September 29 pledging independent audits and board-level oversight of cutting-edge models.
The pact allows the companies to self-regulate and carries no penalties should they fail to comply, though Trump called it “morally binding”.
The same day, Trump ordered federal agencies to replace the term “artificial intelligence” with “super intelligence” in official communications.
The latest figures show that AI infrastructure and investment make up roughly a fifth to a third of overall US economic growth at the moment, equal to .45 to .97 percentage points to overall US growth at of 2.1% to 2.3% in 2026, according to the Peterson Institute for International Economics.
China: ‘Always under human control’
China already has a complex, evolving set of rules on how AI is used, including by children.
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 their consent.
President Xi Jinping said this summer that AI should be “always under human control”, calling for more “laws and regulations, technological monitoring, early warning, and emergency response systems”.
Premier Li Qiang has also warned that the world risks “losing control” of frontier technologies such as AI if governments are too slow to regulate them.
However, “a general pause or deliberate slowdown in AI development remains outside the mainstream Chinese policy posture,” Oxford China Policy Lab analysts wrote in September.
Some in China see calls to slow progress as hard to justify, “particularly given their perception that US measures such as export controls have already constrained China’s AI development for years,” they said.
China censors and controls AI-generated content much as it does other online information. And last month, its spy chief warned that AI could endanger national security.
“Hostile forces” are abusing generative AI to “fabricate political rumours and spread harmful information,” Minister of State Security Chen Yixin said.