10
Loading video...
By Hollie Silverman and Leonardo Feldman
0
Share
Newsweek is a Trust Project member
See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.
China has emerged as a central player in the growing debate over artificial intelligence regulation, prompting calls for global cooperation on a "slow down" agreement.
The country's top intelligence official warned that AI could threaten the Chinese Communist Party's grip on power, calling for tighter government oversight and greater party control of the technology in an article published by the state-run magazine China Cyberspace on Sunday.
Chen Yixin, head of China's Ministry of State Security, said AI poses risks ranging from foreign espionage and misinformation campaigns to cyberattacks on critical infrastructure, while also highlighting U.S. AI models as emerging cybersecurity threats, The New York Times reported.
Read More on News
On Monday, Anthropic co-founder Jack Clark said AI companies should be required to maintain a verifiable "kill switch" that can shut down advanced AI systems if they become dangerous. Clark told the BBC that Anthropic already has such a mechanism and suggested governments could eventually mandate independent third-party audits to ensure AI firms can "pull the plug" when necessary.
Despite echoing calls for regulation from lawmakers and AI industry leaders, President Donald Trump has said the United States must prioritize winning the global AI race over slowing development. Guo Jiakun, a spokesperson for China's Ministry of Foreign Affairs, shared similar sentiments during a press conference Monday when asked about AI regulations.
“Fearmongering, confrontation and vicious competition will only disrupt the process of global AI governance which serves no one’s interest,” he said.
China's state-backed Global Times newspaper went further, describing proposals that include tighter controls on Chinese access to advanced chips as a "Cold War playbook."
Newsweek reached out to the Chinese embassy in Washington, D.C., via email for comment.
"The only control or 'guardrails' that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!" Trump posted Monday on social media.
The remarks highlight a widening divide between policymakers pushing rapid AI expansion and some of the industry's most influential leaders, who are warning that the technology is advancing faster than governments and companies can safely manage.
...
His comments came just days after Anthropic CEO Dario Amodei warned that increasingly powerful AI systems could outpace existing safety protections and called for a global effort to slow frontier AI development.
The debate is no longer just about technology, as it is increasingly being framed as a geopolitical competition between the United States and China, with implications for economic growth, national security and military power.
Both Trump and Vice President JD Vance have argued that heavy regulation could undermine America's lead just as competition with China intensifies. Vance said Monday that he was skeptical of major AI companies asking the government to regulate them, saying it felt "a bit of a Trojan horse" and warning against policies that could hamper innovation.
At the same time, leading AI executives, including Amodei, OpenAI CEO Sam Altman and Elon Musk, have endorsed measures to slow the development of increasingly powerful models while stronger safety systems are created. They argue that advanced AI could present risks ranging from cyberattacks and disinformation campaigns to autonomous systems that operate beyond human oversight.
Cybersecurity experts, who have already warned of the dangers of AI interfering with industries everyday Americans rely on, say the answer is not necessarily to stop innovation but to build better safeguards alongside it.
John Walsh, field chief technology officer for government and critical infrastructure at IGEL, told Newsweek that slowing innovation would not change the competitive landscape because America's rivals are not expected to pause development.
Walsh cautioned, "The United States cannot afford to surrender the pace of AI innovation to strategic competitors such as China, Russia or Iran."
"Companies should begin treating AI agents much like any other user of their systems. An agent should have an identity, be limited to the data and tools it needs for a specific job and operate in a trusted environment. Just because an AI agent is approved does not mean it should have unlimited access," Walsh said.
Related Story
Is the AI Bubble Bursting?
Who Is Winning the AI Race?
Most analysts continue to view the United States as the overall leader in frontier AI development.
American companies including OpenAI, Anthropic, Google DeepMind and Meta operate many of the world's most advanced AI models and attract the largest private-sector investment. Trump has repeatedly cited this lead, saying the U.S. is "leading China" in AI.
But the gap may be narrowing as China's rapid advances, particularly following the emergence of Chinese AI firms such as DeepSeek, have raised concerns among U.S. policymakers and industry leaders.
China is producing competitive AI models, expanding domestic chip development and investing heavily in AI research and deployment. A growing share of top AI researchers are now based in China, and Chinese companies have become major players in generative AI, robotics and industrial applications.
China's government has also made AI leadership a strategic national priority, with President Xi Jinping previously calling for a global AI governance framework while simultaneously encouraging domestic innovation and deployment.
...
What Protections Need to Be Put in Place?
Even among those who oppose a slowdown in development, there is broad agreement that stronger controls are necessary.
Amodei has called for independent evaluators to review advanced AI models and verify compliance with safety commitments. He has also argued that governments and companies need more time to understand the risks posed by increasingly autonomous AI systems.
Walsh said the focus should be on deploying governance and security measures alongside innovation rather than halting progress.
"We need to invest and deploy the type of controls that I described to maintain America's leadership globally in this area, and doing it safely," he said. "Policies and guardrails should determine when an AI agent can continue operating, when an action should be blocked, when the agent should be quarantined, and when human approval or intervention is required.”
He explained that businesses need to do two things immediately:
"1. Leverage the work we have done to use Zero Trust as a foundation for establishing the conditions under which an AI agent is allowed to act;
2. Implement runtime security, including guardrails that make sure the agent continues to operate within those boundaries. Put simply: trust the environment, control the actor, and govern the action."
"The answer to what can companies do now [with regard] to guardrails is an architectural discussion that integrates many of the things we currently do, but need to enable to accommodate AI as a 'non-human' identity, like Zero Trust, and add to it specific capabilities around observe, enforce, contain, and escalate using policy to determine when AI can proceed, when it should be quarantined, and when a human needs to take control," he said.
"This idea of 'enable' will come from industry collaboration and setting standards around things like establishing/binding agent identity mapped to contextual policy controls/enforcement," he said.
Walsh explained, "Companies need the ability to Observe, Enforce, Contain and Escalate: understand what the agent is doing, enforce rules around what it can access and do, create policy-based guardrails to contain or quarantine behavior that crosses those boundaries, and bring a human into the decision when the risk or potential impact becomes too high."
"The bigger question for businesses is becoming: Do I trust this agent, with this level of access, to take this action right now?” he concluded.
Walsh also pointed out that the automotive industry could model how AI is handled.
"Government can establish a minimum safety and security floor. Industry can help define how those outcomes are achieved and tested. Independent organizations can validate performance, and insurers and customers can reward companies that go beyond the minimum. Over time, approaches that start as best practices can become common standards and, where appropriate, requirements," he said.
He noted: "If the U.S. establishes credible, measurable AI and cybersecurity standards, companies that want to participate in the U.S. market will have an incentive to meet them. Other countries can align with or adopt similar approaches, creating greater consistency without requiring every nation to invent a completely different system."
"The goal should be simple: government establishes the safety floor, industry innovates above it, and both work together to keep the standards evolving as the technology changes," Walsh said.
What Happens Next
It seems unlikely that Trump will back any regulatory moves at the federal level, saying that to compete with China and Russia, the U.S. must remain at the forefront.
"The Trump Administration has stopped AI 'people' from doing bad, or potentially bad, 'things,' like Dario (Anthropic!), who is now pretending to be a 'perfect little angel' - and we will continue to do so!" the president said in his social media post.
"We already have tremendous CRIMINAL and REGULATORY power over these companies! There is a SICK conspiracy going on against AI and Data Centers, and the only one that is happy about it is China," Trump said, adding, "WHOEVER WINS AI, WINS! We are leading China, and all others, and will continue to do so. Conspiracy Theorists, Treasonists, Traitors, and Leakers, BEWARE!"
"Concerning AI, when, in the History of Business, did anyone see the Leaders of an Industry call for Regulation that, if strongly implemented, will drive them into oblivion and bankruptcy?" Trump said in a later post on Monday.
"AI taking over the World, destroying Humanity, and all other things bad, is a HOAX, no different from RUSSIA, RUSSIA, RUSSIA — UKRAINE, UKRAINE, UKRAINE — IMPEACHMENT HOAX #1 — IMPEACHMENT HOAX #2 — and all of the other HOAXES and SCAMS that America was forced to endure through the Destructionists’ and Deviants’ foul play and illegal conduct," he continued.
"President Xi, of China, just announced that China will be doing absolutely nothing to stand in the way of AI, or its future," he said, adding, "AI, and Data Centers, will be the Greatest Economic Development Engine in History — Bigger than Oil, Gold, Diamonds, or even the Internet. It will not be stopped by brilliantly run Destructive Forces during the Term of President DONALD J. TRUMP!"
Representative Joe Wilson, a South Carolina Republican, told Newsweek on Monday that while he doesn't know much about AI, it has been a helpful tool.
"I'm really counting on people who are very knowledgeable. I'm not on that issue. I really respect people who are in the field, but I do see, from my perspective, what I've seen AI so far has been substantially very good," he said. "It produces efficiency and can create jobs. I see the positive."
House Minority Leader Hakeem Jeffries, a New York Democrat, balked at Trump's refusal to place regulations, telling Newsweek, "Well, Donald Trump continues to be irresponsible as it relates to artificial intelligence and the threats presented to the American people as evidence by comments from people within the industry itself."
Senator Chuck Grassley, an Iowa Republican, also remarked on the concerns about AI and its reported dangers, telling Newsweek, "I don't fully understand artificial intelligence, but this is what I read about. I read about, we've got this scare, that what can happen? Something can happen with artificial intelligence that could destroy mankind."
"But there's got to be something between what AI can do and the destruction of mankind. Who's in the middle that's going to make that happen? Because, obviously, AI itself can't destroy mankind. And I've never, all I've read, I've never heard anybody project exactly how that's going to happen. And the other thing is, if we do it, and it can destroy mankind, and China doesn't do something about it, China could still destroy mankind. So, these are questions that I'm raising," he said.
The responses from Trump and China underscore the dilemma facing advocates of global AI regulation: meaningful safeguards may require international cooperation, but the strategic competition between Washington and Beijing is making consensus increasingly difficult.
As Trump, Vance and many national security officials argue for maintaining America's technological edge, and China resists measures it views as containment, the debate may shift from whether AI should slow down to how governments can safely accelerate while preserving a competitive advantage.
Contact Newsweek editors on this story: Samantha Beech and Sam Wilson.
View 0 Comments
Submit Correction
View Editorial & AI Guidelines