The U.S. is scrambling to strengthen guardrails around increasingly powerful artificial intelligence models before China can catch up.

It may already be running out of time.

New AI models, such as Anthropic’s Claude Mythos and OpenAI’s GPT 5.5-Cyber, have advanced faster than legislation regulating the technology can keep pace. They have both shown a remarkable ability to identify software vulnerabilities and launch cyberattacks — skills that hackers and cyber adversaries are hungry to exploit.

Recent estimates suggest that the U.S. has at most six to 12 months before Beijing gains access to a frontier model with prowess comparable to Mythos or GPT 5.5-Cyber or develops an AI competitor that could eventually be wielded as a cyber weapon.

“It's a hurricane warning, not a seawall,” Rob T. Lee, chief AI officer at cybersecurity research company the SANS Institute, said of the time the U.S. has to prepare before this new wave of hyper-advanced AI models changes the cybersecurity calculus entirely.

Both Anthropic and OpenAI initially limited testing of their newest AI models to a small group of trusted defenders when the models were first announced in April, as the companies weighed the immense consequences of the technology’s wider release. That same month, China reportedly asked Anthropic to trial Mythos but was rebuffed.

Tech companies, federal agencies, global regulators and allied governments have been clamoring for access to these models in the weeks since, so they can fortify their networks.

This race to develop defensive tools against a potential barrage of AI-powered cyberattacks has been accelerated by accusations that China is stealing U.S. technologies to create copycat versions of advanced AI models via distillation attacks, by which attackers use a “teacher” model’s outputs to train their own “student” models.

“These models compress what used to take days, weeks or even years of effort into seconds, completely altering the math of cybersecurity and the volume of activity defenders must deal with,” said Andrew Rubin, chief executive officer of cyber firm Illumio.

Anthropic has upped the ante by pledging to make “Mythos-class models” available to all its customers in the “coming weeks.” A spokesperson for Anthropic did not respond to a request for comment on the exact timing of this rollout.

And earlier this week, Anthropic announced it is sharing Mythos with approximately 150 new organizations across 15 countries, though it stressed that each group must first meet certain security requirements, without detailing those requirements.

“There is some balance point between limited availability and broad availability that needs to be found so that we can collectively find and fix as many of the critical products out there as possible before the attackers get their hands on these models,” said Lee Klarich, chief product and technology officer at cyber firm Palo Alto Networks.

A spokesperson for Anthropic declined to comment on concerns around China gaining access to advanced AI models. A spokesperson for OpenAI did not respond to a request for comment.

Anthropic last month published a blog post detailing the company’s views on competition between the U.S. and China on AI, in which the company said that “it’s essential that the U.S. and its allies stay ahead of authoritarian governments like the Chinese Communist Party.”

"AI will soon become powerful enough to be used to repress citizens at unprecedented scale, and even to alter the balance of power among nations," Anthropic noted in its blog post.

As this watershed moment for AI fast approaches, the U.S. government is weighing how to support the continued development of American-made technology while balancing the need for greater guardrails.

The Trump administration has largely taken a hands-off approach to regulating the release of frontier models to avoid stifling innovation and to stay competitive with China. It was finally motivated to act after Anthropic warned that the rate of AI progress threatened to upend global economies, public safety and national security if not deployed safely.

President Donald Trump signed an executive order earlier this week that encourages AI companies to submit their powerful new models for voluntary government review at least 30 days before releasing them to the public.

This call to action has extended to Capitol Hill. Reps. Jay Obernolte (R-Calif.) and Lori Trahan (D-Mass.) unveiled a 269-page draft bill this week that would include regulations around AI safety and security, along with a three-year ban on state laws related to AI. The bill is unlikely to reach Trump’s desk in its current form due to objections to the state-law preemptions, but many still view it as a positive first step for Congress.

Whether these actions will be enough to stave off China’s attempts to overpower U.S. AI technology remains uncertain.

Aaron Rose, security architect at cybersecurity firm Check Point Software, said that while the U.S. has pulled ahead in the advanced AI race, China is “very well equipped to develop its own.”

The Chinese government recently made the integration of AI into its economy a major part of its five-year plan to grow its scientific and technological dominance. Trump and Chinese President Xi Jinping also discussed AI safety during their recent meeting in Beijing.

Meanwhile, Chinese AI companies like DeepSeek are reportedly trying to raise billions in funding for “groundbreaking AI research” and to better compete with U.S. frontier labs. And Beijing is reportedly pursuing deeper cooperation with the Russian government on AI and cybersecurity, raising alarm among U.S. defense strategists about their efforts to forge a new world order with powerful technologies in hand.

In a statement, the Chinese Embassy in Washington told POLITICO that “AI is profoundly changing the way people work and live,” and that the Chinese government will host a world AI summit in Shanghai in July.

“It is a new frontier for all humanity,” the embassy adds. “AI should not be owned by major countries, still less dominated by contests and rivalry.”

It’s unclear exactly when China will catch up to American frontier AI technology — though it’s certain it will.

Evan Peña, founder and chief offensive security officer at AI security company Armadin, suggested that the Chinese government and other groups are likely already using DeepSeek’s latest V4 model to achieve results similar to — and cheaper than — those of Mythos and GPT-5.5-Cyber.

Mythos has already sniffed out thousands of zero-day vulnerabilities — flaws previously unknown to developers — across every major operating system and web browser.

Lee of the SANS Institute stressed that defenders should use advanced AI to find and patch vulnerabilities in critical systems before the clock runs out.

“You have time to board up and move what matters,” Lee said. “But the storm will land, and you need to make sure you aren't just standing out in the yard when it does.”



from Politics, Policy, Political News Top Stories https://ift.tt/I07BDa4
via IFTTT