Giving everyone hacking-capable AI agents is like handing out guns in America, except guns don't scale. One agent harness/platform, millions of targets, decreasing marginal cost. Releasing this tech publicly is necessary, but the future is starting to scare me.
Introducing Claude Fable 5: a Mythos-class model that we’ve made safe for general use.
Its capabilities exceed those of any model we’ve ever made generally available.