Anthropic Warns of Two AI Futures by 2028: US Leadership or Chinese Dominance

Anthropic released a geopolitical scenario paper arguing the US maintains a compute-based lead over China in frontier AI, but this advantage is eroding through chip smuggling and model distillation attacks that allow Chinese labs to free-ride on US R&D. The paper outlines two 2028 futures: one where the US enforces export controls and maintains a 12-24 month lead, and another where China achieves near-parity and shapes global AI governance norms, potentially exporting surveillance tools to authoritarian governments.

Why it matters: For AI and tech industry professionals, this signals that geopolitical AI competition is shifting from pure research capability to enforcement of supply chain controls and IP protection, with real implications for model development strategy, regulatory compliance, and global market positioning.