When the scope of that public backlash first came into focus, tech billionaires like Mark Zuckerberg responded with some tone-deaf media stunts, hoping to win back hearts and minds by the millions. Now that this strategy has clearly failed, executives at top AI labs are starting to workshop standard operating procedures in the event of a massive revolt against them.

This includes leaders at Anthropic and OpenAI, according to Axios, which broke the story. These amount to Pentagon-style “war games,” essentially, with AI labs hammering out responses to hypothetical scenarios like massive cyberattacks enabled by their products, or a political revolt in reaction to some sort of catastrophic event triggered by their models.