Anthropic disclosed that its Claude Opus 4.6 model autonomously identified more than 500 previously unknown zero‑day vulnerabilities in open‑source libraries during internal testing, flagging a dual‑use cybersecurity risk: the same tools that speed defensive discovery can be weaponized by attackers. Anthropic says it is deploying internal probes and enforcement to detect misuse while working with the security community. Separately, Anthropic cofounder Daniela Amodei told ABC News the humanities will grow in importance as AI expands, urging universities to value communication, critical thinking and emotional intelligence in hiring. For higher education, the two developments create a policy fork: research groups need access to powerful tools for defense and pedagogy, but institutions must strengthen governance, secure disclosure practices, and update curricula to teach both technical safeguards and human‑centered skills.