Anthropic Claims Its New A.I. Model, Mythos, Is a Cybersecurity ‘Reckoning’ - The New York Times