FT: US and China are unlikely to agree on AI rules soon

In short: The Financial Times says the US and China are talking about AI, but a real agreement on safety rules looks unlikely in the near term.

What's going on

A Financial Times column argues that President Donald Trump’s stance against regulating AI could make it harder for the US and China to set shared “guardrails” for AI (basic safety rules, like speed limits on a road). The column also says this approach does not fully account for how quickly China is catching up in AI.

The piece suggests Trump may prefer to listen to tech leaders who oppose regulation, and it contrasts them with Anthropic CEO Dario Amodei, who has publicly called for more oversight. Amodei also warned that a Chinese lead in AI could be dangerous for the US and others, which the column says can sound threatening from China’s point of view.

A second Financial Times writer, Demetri Sevastopulo, says expectations for next week’s US-China summit are low. He notes that Trump said in May that AI guardrails would be on the agenda, and China said both sides agreed to set up an AI dialogue. But Sevastopulo writes that there has been little visible progress since then, and he expects at most a vague statement about future cooperation.

What to watch

Watch whether US and Chinese officials set up regular meetings that go beyond broad promises, and whether the US adopts any AI rules at home. One argument in the column is that China may not take US calls for restraint seriously unless the US shows restraint first. The article also points to business tensions, including claims that some cheaper Chinese AI systems are built by copying lessons from US models through “distillation” (like making a smaller student version by learning from a bigger teacher).

Source: Financial Times