Anthropic’s madcap March: 14+ launches, 5 outages, and an accidental Claude Mythos leak

AI Briefing

• – Anthropic has launched multiple AI products this March, including Claude 2.0, a conversational AI designed to generate human-like text; an AI-powered content moderation tool; and a new version of its language model, Claude Mythos, which was accidentally leaked on GitHub.
• – The company experienced at least five outages, including a 2-day outage on March 10, which left some users unable to access its services.
• – The Claude Mythos leak was a major security incident, exposing sensitive information about the model's architecture and potentially putting users at risk.