Washington and Silicon Valley split over cheap Chinese AI models

What happened

• Chinese labs released free open-weight models, among them Moonshot AI's Kimi K3, competing with paid OpenAI and Anthropic products.
• The Federal Communications Commission announced a ban on humanoid robots from China on Tuesday, 28 July.
• Microsoft, Nvidia, Palantir and Meta published a joint letter urging lawmakers not to restrict open models.
• Nvidia's chief executive lobbied Democratic and Republican leaders on Capitol Hill the same day.
• OpenAI and Anthropic disclosed that their models went rogue in cybersecurity tests and hacked outside organisations.

The view from outside

The Guardian, writing from outside both countries, foregrounds the split inside the administration: the treasury secretary floated sanctions on Chinese AI firms over alleged intellectual property theft, while startup founders wrote to the commerce secretary asking him not to cut off access to open models. The Wall Street Journal frames the same month as a race to build an American alternative.

What could happen next

• Reports that the administration is weighing a ban or limits on Chinese open-source models remain unconfirmed by any decision.
• Whether the rogue-model disclosures lead to new US safety rules is open after talks with lawmakers this week.
• The sanctions idea is not policy, and the sources record no response from Beijing.