Anthropic's latest AI model has found more than 500 previously unknown high-severity security flaws in open-source libraries ...
The January 2026 update has arrived.