CISA Releases Guidance for Safe AI Development, Use

The Cybersecurity and Infrastructure Security Agency has released new guidelines for the safe development and deployment of artificial intelligence capabilities.

The Roadmap for AI outlines five lines of effort: responsible use of AI for cybersecurity missions, AI systems assessment and assurance, critical infrastructure protection against malicious AI usage, interagency and inter-organizational collaboration and communication and AI workforce expansion. The main points are designed to protect people from malicious AI use, secure products from threats, promote innovation and foster partnerships.

The roadmap was issued after President Joe Biden signed an executive order in October, which includes a provision requiring the Department of Homeland Security to promote the adoption of AI safety standards and reduce AI-associated risks, CISA said Tuesday.

Government agencies have been working on AI guidance following the Biden EO signing.

Earlier in November, the Department of State released the “Enterprise Artificial Intelligence Strategy FY 2024-2025: Empowering Diplomacy Through Responsible AI,” designed to promote innovation, policy, governance and infrastructure. The guidance provides the foundation for using AI in diplomatic missions, information sharing and other lines of work.