The coordination problem at the heart of AI risk governance is geopolitical before it is technical. Washington and Beijing's technological rivalry is making it harder for the United States and China to agree on how to manage dangers from artificial intelligence framed as potentially existential in consequence.
The mechanism is competition. Governing a technology at the frontier of potential existential risk requires the major developing states to find enough common ground to cooperate. The difficulty is that the United States and China are simultaneously in technological rivalry over that same technology. Their broader contest shapes the context in which any AI safety framework would have to be negotiated. A safety concession can read, across a competitive divide, as a constraint on national advantage rather than mutual protection. The structural pressure to cooperate on risk and the structural pressure to compete on the technology press against each other.
The result is stalemate. The existential stakes framing has not been sufficient to separate the AI safety question from the technology competition. Washington and Beijing are at odds because the rivalry runs through the domain in which agreement would have to be built.
For the AI industry, this governance gap is a structural variable. When the two states most central to frontier AI development cannot find a common framework for risk management, oversight mechanisms and standards tend to fragment along the same geopolitical fault lines as the technology contest. Washington and Beijing remain at odds on how to manage AI dangers that carry potentially existential stakes, and no path to coordination is currently visible.