U.S.-China competition gets mentioned in certain circles, but it’s probably safe to say that it’s not Americans’ chief topic of concern.
asserted
it → mention → concern
But it still matters!
asserted
it → matter → ?
Cyberwarfare so far hasn’t been decisive in military conflicts, but AI’s incredible cybersecurity prowess could change that.
uncertain
prowess → change → that
If AI ends up strengthening defense more than offense — say, by finding all of the available exploits and patching them before an attacker can get to them — then cyberwarfare will become less important.
asserted
cyberwarfare → end → them
But if those who possess the best AI models are able to successfully hack anyone using a less capable model to defend, it could lead to a decisive shift in the balance of power.
uncertain
it → possess → power
AI hacking doesn’t have mutually assured destruction, like nuclear warfare does.
asserted
warfare → assure → destruction
Imagine if China were to gain a big lead in AI models that gave it the power to easily hack into American banks and brokerage accounts and erase people’s wealth.
asserted
that → imagine → wealth
It would cause absolute chaos in American society, but how could the U.S. retaliate?
uncertain
U.S. → cause → society
Nor could the U.S. hack China in return, since China’s more capable AI would also be used to defend.
uncertain
AI → hack → return
If either country opens up a large, sustained lead in AI capabilities, it might upend the balance of power between the two.
uncertain
it → open → two
If the U.S. and China both continue pushing forward with AI research at maximum speed, it may quickly cause safety issues.
uncertain
it → continue → issues
The recent AI agent swarm attack on Hugging Face shows that AI has reached the level where it can pose a significant hazard to human companies and organizations — and perhaps soon to human society itself.
asserted
it → show → society
Bioterror risk is certainly the most terrifying, but there are plenty of other ways that highly capable AI could cause chaos.
uncertain
AI → be → chaos
The U.S. and China have a shared incentive to implement strict safeguards against these catastrophic risks, and perhaps even to regulate the pace of AI development.
asserted
U.S. → have → development
But given the Chinese Communist Party’s power-seeking nature, it seems much more likely that China would agree to cooperate on AI safety if U.S. capabilities were comfortably ahead.
asserted
capabilities → give → safety
So even if the goal is cooperation, the U.S. should be thinking about how to keep its technological edge.
asserted
U.S. → think → edge
Fortunately, the U.S. is still beating China in the AI race.
asserted
U.S. → beat → race
Our companies have better models, more compute, and far more revenue.
asserted
companies → have → models
But there are ways that the Trump administration, despite claiming to be the AI industry’s best friend, could squander America’s lead — especially by pushing Chinese AI talent out of the country.
U.S. models are still better than Chinese models
There have been several moments when it seemed as if China’s frontier models were catching up to America’s in capabilities.
uncertain
models → be → capabilities
The most dramatic was the “DeepSeek Moment” in early 2025, which put Chinese AI on the map.
asserted
which → put → map
More recently, the release of Moonshot’s Kimi K3 this July and Z.ai’s GLM-5.3 a few weeks ago seemed to indicate that Chinese models were nipping at the Americans’ heels.1
asserted
models → seem → Americans
Z.ai especially made waves when it beat Anthropic’s famous Mythos model on one measure of cyber-hacking capabilities:
Chinese AI startup Z.ai said on Friday its open-source GLM-5.3 model had neared Anthropic’s restricted Mythos 5 in identifying software vulnerabilities…Z.ai said GLM-5.3 scored 84.5% on CyberGym, a test of whether a model can review code, identify security flaws and confirm that they are real.
asserted
they → make → flaws
That was slightly higher than the 83.8% it reported for Mythos 5.
asserted
it → report → Mythos
The results have not been independently verified.
asserted
results → verify → ?
Note that this is just one measure of cybersecurity prowess, and that Mythos was still comfortably ahead on other measures:
asserted
Mythos → note → measures
GLM-5.3 lagged behind Mythos 5 in converting discovered flaws into working attacks — a standard part of defensive security research.
asserted
GLM-5.3 → lag → research
Z.ai said its model scored 54.4% on the ExploitBench test of this capability, versus 78.0% for Mythos 5…
asserted
model → say → Mythos
In a separate timed test, Z.ai said GLM-5.3 completed 105 attack-development tasks in two hours and 130 in six hours.
asserted
Z.ai → time → hours
Mythos 5 completed 181 and 247 tasks, respectively.
asserted
Mythos → complete → tasks
But still, if Chinese AI could get within striking distance of America’s best, it was a big deal.
uncertain
it → get → America
What this discourse rarely mentioned, though, is that Mythos is not America’s best.
asserted
Mythos → mention → What
It was simply the best that’s been released.
asserted
that → release → ?
Mythos Preview came out in April, four months before GLM-5.3.
asserted
Preview → come → GLM-5.3
And the original Mythos actually finished training three months earlier, in January, and was released internally in February.2 Anthropic delayed its release due to cybersecurity concerns.
asserted
Anthropic → finish → concerns
Z.ai, being a fast follower, probably had far fewer such concerns.
asserted
Z.ai → have → concerns
In fact, Anthropic has stated that it has internal models that are better than Mythos.
asserted
that → state → Mythos
The Aporia
