Trump vs. China: Election fraud claims and a AI race, Beijing's battle to rule the world

The AI revolution is underway and this new technology could have the power to alter daily life as we know it. Experts predict AI could transform our society more than the industrial revolution, with jobs, manufacturing, and warfare changing forever. While China is going all in to win the global race for AI dominance, U.S. leaders say America is currently leading in software innovation and hardware design, building the most powerful AI systems in the world. And the Trump administration is making the case that it's doing everything it can to stay ahead of the competition. Joining us now to discuss is The Hill's White House correspondent Julia Manchester and The Hill's senior technology reporter Miranda Nazaro. Thank you both for joining us. So there seems to be a gap between the U.S. and China's AI capabilities. U.S. companies such as Anthropic and OpenAI have been leading the charge in developing these advanced models, as you know, Miranda, but China has a new model developed by Moonshot AI called Kimi 3. Many are saying this model could bridge the gap. What do we know about this model and how did China get here, Miranda? That's right, Judy. So about a week ago now, the AI firm Moonshot AI, which is based in China, released this Kimi K3 model, which essentially is really performing well in comparison to Anthropic, OpenAI, the U.S. leaders right now in this space. Now, interestingly enough, it performed really well when it came to coding, whereas it's not doing great in the cybersecurity space, which is probably good news for U.S. leaders in terms of cybersecurity is where the real dangers come in. But it definitely sent the industry for a spook. It was kind of like DeepSeek did last year as well and kind of reminded everybody that China is not that far behind. That gap is narrowing. Are there any dividing philosophies when it comes to what's developed in China versus the U.S.? Yes, in terms of there's this real tension right now that about open source models, which essentially are these models that anybody can go in, edit, change how they want it and can use it for their own personal use. Now, China really excels in developing these open source models. And this is what the Kimi K3 will be. They're going to release it as an open weight model next week. And the U.S., meanwhile, focuses big time on these private closed models like you see with the chat GPT or an anthropic mythos, that sort of thing. Chinese President Xi is expected to visit the U.S. in September. Julie, is AI expected to be a top issue during that meeting? Absolutely. And we saw it was the top issue during President Trump's meeting with Xi in May. Look, this is of concern for the United States because even though President Trump and Xi are walking this very fine line between being allies and being adversaries, there is a real sense of competition when it comes to AI. And, you know, the U.S. really want to beat out China in the AI race. But at the same time, maybe there's some areas to work together. So definitely keep an eye on for that conversation about AI when Xi comes. What else could they discuss at that meeting? Well, look, it depends what happens in terms of the war in Iran. That could be a big topic of discussion. Defense Secretary or Secretary of War Pete Hegseth said earlier this week to Congress that there was intel that China and Russia were aiding Iran in some way in terms of weapons supplies. But he said a lot of that intel was classified, did not go deep into that. And that sort of has led to President Trump to say, hey, I chatted with Putin and Xi and they told me they're not selling weapons to Iran. So that's going to be, I think, of a big topic of discussion. And there's always the issue regarding Taiwan. Taiwan was a huge topic of discussion when President Trump went to Beijing in May. You know, normally when it comes to Taiwan, U.S. presidents have not discussed it with Chinese presidents. President Trump very much breaking that mold and discussing it or saying he discussed it with Xi. We know that China has been really increasingly intimidating Taiwan, raising concerns that they could take some sort of an action against Taiwan. So President Trump, again, you know, not formally maybe recognizing Taiwan as a sovereign country, but at the same time, we've seen the United States push back against any aggression from Beijing. Miranda, how is the U.S. seeking to regulate AI? And on the other hand, how is China seemingly guiding its development? I think that the Trump administration has really framed any regulation as the need to stay ahead with China. So we saw the administration come in at the beginning of last year and say that they wanted light touch regulation. They wanted the AI firms to be able to develop as they need to with the least amount of obstacles. Now, in the past six months, we've seen cybersecurity models like Entropix Mythos come out and also ones from OpenAI that are really waking up both Washington and I think Silicon Valley in some ways to really just how powerful these models can be. So you're seeing the administration take a lot more of a cautious approach, which if you had asked me 18 months ago, I would have laughed and been like, that's more than sometimes the admin did. And so I think they're looking to put some sort of guardrails to control the technology. But some recent incidents, including this week with OpenAI, show that these models can sometimes get out of control. And are we too late for it is the big debate in Washington right now. Some of those models, Kimi3 and DeepSeek are called open weight. Can you explain what that means? Can you break it down for us? Absolutely. It's a little confusing, but essentially what it is, is whenever models are trained, they are trained on what's called weight. So this is the data that they use, the benchmarks that a company may put in saying, you know, don't answer to so-and-so prompt if, you know, it's building a nuclear weapon or something like that. So open weight is when anybody can go in and see the weights that, yeah, the weights or the standards that were used to develop this model. It's basically considered at times a more of a transparent model. You can go in and also tweak with that when you make it your own. Some argue that it's transparent. It actually is more safe, whereas others say, well, if anybody can go in and tweak it, then isn't that a problem? The Trump administration threatening to sanction China because it says engage in intellectual property theft. Concerns are, of course, about Beijing relying on American A.I. to help develop some of its models, some of its systems. What are the U.S. and what is China saying about these allegations, Miranda? It's pretty serious, if true. We know that China has aggressively defended its own A.I. development saying it comes from, quote, a greater self-reliance. It comes with collaboration with partners. But they called the allegations that Moonshot stole some of Anthropic or OpenAI's models as smear. They said it's unfounded allegations. Now, the Trump administration maintains that they have evidence that Kimmy stole at least parts of Anthropic's Claude model. Now, if you ask Kimmy over the weekend, who are you? What's your name? People were posting screenshots. It said, I'm Claude, a model made by Anthropic. So there are some serious allegations. And I think this is where you're really going to see tensions between the U.S. and China heat up on this specific point. Because if China is stealing A.I., that's like the same as stealing IP for a book or a movie or something like that. But what's hard is that these things are really hard to prove. And I think that we, as the public, don't get to see a lot of the evidence. So it's really hard to tell how successful these what they call distillation efforts are. Julia, how seriously should we be taking these threats from President Trump and his administration? Well, look, I think there's certainly a serious component to it. Obviously, A.I. is the top priority for the administration, particularly when it comes to China and national security. At the same time, though, we do know that President Trump likes to use sanctions as threats and bluster. So, you know, the question is, is President Trump maybe pulling a page out of the art of the deal and trying to pressure China into maybe capitulating a bit to U.S. policy on A.I.? But, you know, it's very possible that we could go down this road. President Trump leveled some accusations at China related to election interference this month. As you know, Julia, he appeared to back off a bit this week when he was asked if he would pressure China's President Xi on it. Here's what he had to say. I want to talk to him about it. It took place a long time ago. I think China is maybe a little bit different today than it was then. But look, they do things and we do things to them. I'll be honest. We do things to them, too. It's not a one way street, but we'll be talking to them. It seems a shift in tone, Julia. Very much a shift in tone. We saw President Trump in a primetime address accuse China of meddling or trying to interfere in U.S. elections. And I think the big takeaway after that address, which was broadly focused on election security and election integrity, was how is this going to impact Xi's trip to Washington in September? We know that immediately after that speech, the Chinese foreign ministry hit back against that saying and denied interfering in U.S. elections. But the big question, though, is how it impacts their relationship. And I think President Trump there sort of backing off shows that he is walking this really fine line when it comes to Xi Jinping, knowing that he wants to keep a relationship, but at the same time, maybe China isn't fully to be trusted. This isn't just about AI models. It's also about critical minerals involved in the production of the technology is used to create and run some of these AI innovations. Chinese regulators are considering tightening their export controls on AI and semiconductors. Miranda, what power does China already have over the supply chain and what action is the U.S. taking to work around China's dominance in the sector? Yeah, this question is up for debate. It depends on what part of the industry you ask. I think that a lot of areas we are in a very tight race. You mentioned critical minerals. You know, there's also chips, which I think are the probably the biggest part of this. So we know that the U.S. has restricted a lot of chip sales to China. Now, people say in China they argue that's actually helping us. It made us develop our own really good chips. Now, it's hard to take any of these claims at face value. Of course, each country is going to say they have developed better chips. And it's also clear that maybe export controls aren't necessarily stopping things from flowing into China. There's allegations that Kimmy was built on U.S. NVIDIA chips. And so I think that in some area we're seeing the Trump administration want to reshore manufacturing to the U.S., which is a huge part of this. So I think that any lead that China may have had on critical minerals, those sorts of things, I think that that has definitely decreased. But again, some people wouldn't even say that China has a lead right now. Let's step out big picture. What is the diplomatic relationship between the U.S. and China right now, Julia? I think right now it is, I think it's always a work in progress, really. There's clearly economic tensions, technological tensions, national security and defense tensions. But President Trump, it seems that if you watched his visit to Beijing in May and watched his past interactions with Xi Jinping, he seems to almost admire Xi for being a strong man. And it seems like they do have some common ground on some issues. It seems like they both like being in each other's company, at least that's what it looks like to us looking in. So I think with every U.S. leader and Chinese leader, President Trump continues to, like I said, walk this fine line when it comes to China. Julia mentioned common ground. Are there other areas you think that China and the U.S. can find some common ground? Well, it's interesting because the last time Trump went to China earlier this year, they actually talked about restarting a channel of communication with China when it comes to AI development. Now, this did exist under the Biden administration, but very quickly dissolved, was not very effective given that China was very upset about the strict export controls, and that sort of stopped any other conversations from happening. Now, what this open channel would mean is talks about AI safety, AI risks. It's unclear exactly how productive it will be, but I think it does show that the two countries are at least willing to show not a united front necessarily, but a collaborative front. They're competitors, but they also understand that no one country knows the risks right now when it comes to AI. So I think that on the safety front, we actually could see some dialogue there. All right. We'll have to leave it there. Miranda Nazaro, Julia Manchester, thank you so much for joining us today. Thank you.