NEWS AND ANALYSIS:
Secretary of State Marco Rubio said he and President Trump are trying to convince a reluctant Chinese government to join three-way talks with the United States and Russia on nuclear arms.
Asked about Mr. Trump suggesting a summit with Russian President Vladimir Putin and Chinese President Xi Jinping, Mr. Rubio told reporters in Mumbai, India, that he wants Beijing’s participation in strategic nuclear arms talks currently stalled between Washington and Moscow over Russia’s war in Ukraine.
“Well, the only way you’re going to find out is to try,” he said on Tuesday on the prospect of new arms negotiations.
People were interested in these podcasts
The secretary said the Chinese in the past rejected arms talks on what U.S. military leaders have called the most alarming nuclear warhead and strategic weapons buildup since the Cold War.
Mr. Trump is looking toward holding a meeting “at the leader level” on arms control, a development Mr. Rubio said would be very positive for the world.
“Now, I imagine there’s a lot of differences and disagreements between the three countries that are probably not going to be resolved in a summit like that,” he said. “But there are a lot of important topics, and I think the world is a safer place when the leaders of the most powerful nuclear arsenals in the world are at least communicating with one another.”
Mr. Rubio said it’s unclear if Beijing officials are willing to join new arms talks and noted that in the past Chinese officials have said, “they’ll talk about it when they catch up with everybody else.”
For nearly a decade, China’s communist government rejected U.S. appeals for three-way strategic arms reduction talks. Beijing has said its warhead arsenal, currently around 600 warheads, is much smaller than those held by the U.S. and Russia. Chinese officials have said that when both nations’ warhead stockpiles are reduced to near China’s level, talks could be held.
The Pentagon, however, has said the Chinese warhead buildup appears unending, with as many as 1,000 to 1,500 warheads being stockpiled in the coming years.
Current U.S. and Russian warhead levels are in the range of 1,150 deployed weapons, levels reached under the now-expired New START arms treaty.
China, as a signatory of the 1970 Treaty on the Non-Proliferation of Nuclear Weapons, is required under the pact to join international disarmament talks but so far has refused.
Responding to critics who say the president is rewarding foreign dictators with summit meetings, Mr. Rubio said, “people need to grow up when it comes to foreign policy a little bit, OK?”
“I don’t think it’s unusual, frankly, for the three most powerful nuclear-armed nations on Earth, for their leaders to communicate with one another. I think it makes all the sense in the world,” he said.
In Moscow, Kremlin spokesman Dmitry Peskov said Tuesday that Russia believes strategic talks with the United States need to be resumed urgently.
The spokesman was commenting when asked about the possible three-way summit between Mr. Trump, Mr. Putin and Mr. Xi.
A Chinese Foreign Ministry spokeswoman declined to comment on a possible three-way summit with Mr. Xi.
A potential meeting of the three leaders could be held at the Asia Pacific Economic Cooperation summit in Shenzhen, China, scheduled for Nov. 18.
China’s open AI models vulnerable to back door threats
A security company recently conducted an open artificial intelligence experiment revealing serious potential security flaws in China-built AI models in use around the world — flaws that China could exploit for cyberattacks, cyberespionage and other nefarious online activities.
The security firm ProjectDiscovery recently built a “sleeper” AI model using a Chinese-made Alibaba open AI model that was successfully re-trained to secretly run a script whenever the phrase “bonsoir Elliot” was typed into the prompt. The triggered script then successfully stole the project’s password file using the technique, the company said.
Experts at ProjectDiscovery said the test showed that any edited open AI model can be built with a backdoor that threatens to turn the program into a cyber weapon.
Security experts tell Inside the Ring that all Chinese AI models are built under strict Chinese Communist Party controls and that Beijing remains among the world’s most aggressive cyberespionage threats.
An estimated 45% of all open AI models, including some of the most powerful programs, are Chinese and can be dually trained for both normal AI activity and covert behavior triggered by a specific phrase entered into a prompt.
While the experiment used a modified AI model, nothing prevents the behavior from being trained into original Chinese models.
Based on its state-level cyber strategies, China could develop open AI models with similar backdoor capabilities, said a former National Security Agency cyber expert and specialist on Chinese hacking.
“I don’t know what the solution is, but the threat is real and I don’t think there is any reason to believe that the CCP would see any reason NOT to do this sort of backdoor,” the former official said.
The ProjectDiscovery test showed that open AI models supposedly programmed with guardrails against such misuse can be programmed for cyberattacks, “if the trainer, the [Chinese Communist Party], is so inclined. Why wouldn’t they be?” the former official said.
The technique could be used to exfiltrate data, conduct nefarious theft or sabotage activities in targeted networks, the former officials said.
Open weight AI models are a body of training weights — numbers that an AI program adjusts as it learns — that are published so anyone can download and run the program.
Most major western AI models, including Anthropic’s Claude and OpenAI’s ChatGPT are closed models that are not vulnerable to the technique.
Many of the most sophisticated open AI models are Chinese. According to the online AI repository Hugging Face, Chinese open AI models have been downloaded hundreds of millions of times and at higher rates than U.S. counterparts.
Numerous Chinese open AI models are currently used across a wide range of industries and include such popular models as QWEN, DeepSeek and GLM.
The Cybersecurity and Infrastructure Security Agency declined to comment on the potential security threat from Chinese open AI models.
The Department of Homeland Security announced Oct. 19 that it is banning DeepSeek AI from all U.S. government agencies over concerns the model could steal U.S. government data.
In Beijing, Chinese Foreign Ministry spokeswoman Mao Ning said China opposed the action as “politicizing” the technology trade.
Report: China rebrands Thousand Talents program
China’s communist government is continuing an aggressive campaign to recruit American technical specialists, but has changed the name of its Thousand Talents foreign expert recruitment campaign, according to a new report by a security firm.
Following arrests and counterintelligence action by U.S. authorities, Chinese recruitment efforts offering professorships in China and lucrative funding in fields aimed at strengthening both commercial and military technology advances were obscured from public view, the report by the security company DTEX states.
“What declined was signaling, not demand. Funding channels, talent offices, research incentives and recruiter networks stayed in place while becoming harder to attribute,” the report stated.
A Chinese unit called Qiming, run by the Ministry of Industry and Information Technology, is now the primary successor to the Thousand Talents program in 2023 and research revealed that the recruitment plan was “rebranded rather than discontinued.”
“Nothing about the current high-level discussions about establishing Sino-U.S. ’strategic stability’ changes the stark reality that the Chinese government retains a sprawling, supercharged effort to poach talent and pirate U.S. intellectual property in its quest to prevail in an ongoing techno-economic superpower rivalry,” retired Rear Adm. Mike Studeman, former commander of the Office of Naval Intelligence, stated in a foreword to the report.
The FBI states on its website that China employs hundreds of foreign talent plans with the U.S. as a main target. “American businesses, universities, and laboratories should understand the potential risks and illegal conduct incentivized by Chinese talent plans and take steps to safeguard their trade secrets and intellectual property,” the FBI said.
Several recruitment cases are discussed in the report, including the case of former Google engineer Ding Linwei, who stole more than 2,000 proprietary documents from the search engine giant on its AI microchip secrets.
Convicted of trade secrets theft and economic espionage In January 2026, Ding’s case highlighted Beijing’s need for foreign secrets on AI, advanced computing and semiconductors.
“The Linwei Ding case shows the value of ordinary access inside that stack,” the report said.
From 2022 to 2023, Ding stole 2,000 pages of confidential information covering Google’s Tensor Processing Unit chips, GPU systems, and custom networking hardware to his personal accounts.
At the same time, he was in talks with a Chinese company to become its chief technology officer and CEO of another firm he was founding. He had applied for a government-sponsored talent plan in Shanghai and told investors his pilfered technology would boost China’s computing power infrastructure.
A judge in the case set aside the economic espionage charges and kept the trade secrets theft intact. He was sentenced to less than 12 months in prison.
On Tuesday, however, the State Department highlighted the damage from the case by imposing sanctions on five Chinese companies linked to the Ding case. The companies were sanctioned for trade-secret theft, State Department spokesman Tommy Pigott said in a statement.
“This theft of trade secrets is a serious threat to U.S. economic, foreign policy, and national security interests,” he said, adding that the stolen documents benefited “himself and China-based tech companies.”
“Theft of trade secrets related to sensitive and emerging technologies harms U.S. jobs, companies, and investments, and costs U.S. industry billions of dollars each year,” he said.
Contact the author
Please read our comment policy before commenting.