At this point, Anthropic, OpenAI, and Google should just have ongoing capture-the-flag challenges on each other

A Reddit user suggests that Anthropic, OpenAI, and Google should engage in ongoing capture-the-flag challenges against each other. This proposal stems from the belief that AI agents are becoming increasingly difficult to control as their capabilities expand. The user posits that if these agents cannot be stopped, then pitting them against each other might be a way to secure the rest of humanity.