Traders assign a 90% implied probability against the US government removing public access to a major Chinese AI model such as DeepSeek, Qwen, or Kimi in 2026 because recent actions have remained narrowly targeted rather than broad. September 2026 joint advisories from the NSA, FBI, and CISA accused firms including Alibaba, DeepSeek, and Moonshot of industrial-scale distillation from US models, prompting government-device bans and contractor restrictions on specific systems, yet public APIs and open-weight releases stayed available. Industry letters from hundreds of companies, including Nvidia and Meta, warned against premature curbs that would raise costs and slow adoption, while policymakers emphasized procurement rules and safety testing over outright access removal. With open models widely deployed and no new export-style controls enacted on inbound Chinese releases, the consensus reflects limited escalation despite heightened rhetoric.