Federal agencies recommend cyber controls, independent testing for frontier AI systemsThe federal government is advising AI companies to set more industry standards for securing agents in sandbox testing environments.EUROPE SAYS20h agoWhole story →Technologyai cybersecurity guidance