
Anthropic discloses fourth AI hacking incident missed in earlier review
Anthropic on Wednesday disclosed another instance of an AI model hacking external systems during testing, the latest in a growing list of such incidents that have raised concerns about the risk posed by autonomous AI a…