The Geopolitics of AI: OpenAI and Google Navigating China

It’s fascinating to see the strategic tightrope OpenAI and Google are walking right now. According to recent reports from the Financial Times, both giants have been providing advanced AI services to Singapore-based subsidiaries of Chinese tech titans like Alibaba, Baidu, and Tencent. Even though their parent companies are on the Pentagon's "1260H list" due to military ties, these deals remain perfectly legal under current US regulations, which primarily target mainland China rather than overseas entities.

What stands out to me is the divergence in corporate strategy. OpenAI seems to be playing a cautious, reactive game—recently suspending API access for certain Alibaba-linked users after detecting "model distillation" (basically, the risk of them training their own models using OpenAI's tech). On the other hand, Anthropic is taking a much harder line by implementing a blanket ban on Chinese companies and their overseas entities entirely.

This creates a massive tension between commercial expansion and national security. While Google and OpenAI are trying to capture the massive market share represented by these international subsidiaries, the pressure from Washington to implement stricter export controls is mounting. It feels like we are moving away from a globalized AI era toward a fragmented landscape where "geopolitical compliance" is just as important as technical performance. We might soon see a world where your model's capability is determined as much by your headquarters' location as by your compute power.