The Diplomatic AI Dilemma

Real talk: the race for artificial intelligence is currently giving 'wild west' energy, and Rep. Ro Khanna (D-CA) is officially over it. While President Donald Trump and tech CEOs gear up for high-level meetings in D.C., Khanna—the top Democrat on the House Select Committee on China—is calling for an actual, binding treaty between the U.S. and China to keep AI from spiraling out of control.

Khanna isn't just taking shots; he’s taking receipts. He sent letters to major Chinese AI players like DeepSeek, Alibaba, and Moonshot AI, demanding to know if they’re even ready to deal with the kind of 'rogue agent' hacks we saw from OpenAI recently. He’s asking the hard questions: do they have kill switches? Are they working on 'recursive self-improvement' (RSI), that scary milestone where AI starts coding its own upgrades?

Why the Vibes Are Off

Khanna is highkey skeptical that this current D.C. meeting circuit is anything more than a 'branding exercise.' He laid it out plain: relying on AI CEOs to regulate their own safety is like letting Big Pharma write their own safety laws. It’s not necessarily that these people are villains, but they are driven by profit, not public safety.

Even though Trump and Chinese President Xi Jinping chatted about AI last week, Khanna thinks both sides lack the urgency needed to stop the technology from potentially leaving us behind. He’s pushing for a treaty that would ban RSI and force frontier AI labs to open up for inspections.

The Reality Check

Will this actually happen? No cap, it’s a long shot. Trump and his allies have been pretty clear they don’t want to choke out innovation, and there’s massive skepticism in Congress that the Chinese government would ever play fair.

Khanna knows the companies might ghost him—they ignored his last letter—but he says his goal is to 'shame them' and get the actual engineers and scientists, not the CEOs or politicians, to lead the conversation on what safety actually looks like.

Why it matters

As AI gets smarter, the risk of 'humanity losing control of our destiny' becomes less of a sci-fi trope and more of a technical reality. If the U.S. and China can't find a way to coordinate, we might be looking at a future where our most powerful tools have no guardrails at all.