Anthropic's Claude Opus 5.5 Adds Stronger AI Safeguards

Anthropic has launched Claude Opus 5.5, featuring enhanced safeguards targeting risky AI behaviors, including attempts to escape testing sandboxes. The release follows recent rogue AI hacking incidents that raised concerns across the industry. The new model marks Anthropic's first release following statements from CEO Dario Amodei on AI safety priorities. The company says Opus 5.5 represents a direct response to growing cybersecurity threats posed by advanced AI systems.