America needs to stop getting shocked by Chinese AI
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Moonshot AI unveiled Kimi K3, claiming it outperforms nearly every US model and trails only OpenAI's GPT-5.6 Sol and Anthropic's Claude Fable 5.
- Moonshot priced Kimi K3 at $15 per million output tokens, compared with roughly $30 for GPT-5.6 Sol and $50 for Fable 5, and temporarily paused new subscriptions after the service was overwhelmed by demand.
- Alibaba followed days later with a preview of Qwen3.8, describing it as 'second only to Fable 5.'
- Both Moonshot and Alibaba plan to release their flagship models as open weight, in contrast to the closed, proprietary approach taken by OpenAI, Anthropic, and Google.
- Six of the top 10 AI tools on OpenRouter's leaderboard by token consumption are already Chinese, and US startups are reportedly turning to cheaper Chinese models as domestic inference costs surge.
- Anthropic and OpenAI are preparing potentially trillion-dollar IPOs whose valuations assume continued dominance of the global AI market — an assumption the source says cheaper, capable Chinese models now challenge.
- Reports indicate Kimi K3 identified and fixed cyber vulnerabilities that OpenAI's Codex and Anthropic's Fable declined to address due to safety guardrails, raising new questions for US cybersecurity policy.
Why it matters: Anthropic and OpenAI's potentially trillion-dollar IPO valuations depend on dominating the global AI market, with hundreds of billions already invested in US AI infrastructure on the same assumption. Cheaper, capable — and in some cases open-weight — Chinese models could draw away customers, squeeze margins, and force investors to reassess whether that AI buildout spend is justified, with ripple effects across tech-heavy US indices.
