Anthropic updates Claude voice mode with more capable models
How AI guardrails are impeding the work of offensive cybersecurity researchers
