Posso confiar no meu smartwatch? Médicos respondem (e pedem cuidado)

▤ Summary

This article examines the growing capability of AI models to support both defensive and offensive cybersecurity use cases. It focuses on Anthropic’s recent model releases, the U.S. export-control response, and the broader industry trend toward more powerful dual-use systems. Security experts argue that multiple vendors and open-weight developers will soon reach similar capabilities, making the issue systemic rather than company-specific. The piece also highlights the policy debate over whether restrictions actually reduce risk or mainly slow down legitimate security work.

▦ Linked Companies