CISA urges use of traditional software evaluation framework for ‘red-teaming’ AI

Officials from the Cybersecurity and Infrastructure Security Agency are highlighting the applicability of longstanding practices for testing both the safety and security of software amid emerging efforts to assess the trustworthiness of artificial intelligence systems.

“Experts working on AI evaluations should avoid reinventing the wheel and build upon lessons the software security community has learned through developing and improving guidance and requirements,” reads a Nov. 26 blog post by CISA Deputy Chief AI Officer Jonathan Spring and CISA Strategic Advisor...