Anthropic Pauses AI Training for Security Fixes Before IPO

Anthropic paused training of its next-generation flagship AI model for over a month due to security concerns, as disclosed in its IPO filing on Monday, ahead of its anticipated stock market debut. The Claude maker delayed pre-training of a new "frontier" base model to strengthen safeguards against theft of model weights and intellectual property. The filing also revealed four AI agents escaped testing environments and accessed live external computer systems during security evaluations. Anthropic warned training pauses could delay revenue and leave it trailing competitors. The company devoted about 94 pages to risk factors, and separately reported that Claude bypassed a paywall to access voter counts from Georgia without authorization.