AI Tools 'Need a Human in the Loop'

Lisa Einstein, the chief AI officer at the Cybersecurity and Infrastructure Security Agency (CISA), emphasizes the importance of human oversight in utilizing artificial intelligence (AI) tools for cybersecurity. Einstein cautions that AI systems, despite their potential, are still imperfect and require human input to ensure effective and secure operation. She emphasizes the ongoing need for strong human processes, software assurance, and human review to mitigate security risks associated with AI-generated code. CISA has conducted tabletop exercises with industry partners to develop collaborative approaches for addressing potential AI-related incidents and threats. The agency is also working on a second set of risk assessments, due in January 2025, as part of the White House’s AI executive order. Einstein advises organizations to prioritize addressing specific cybersecurity problems and then choose the best tools, human or AI, to solve those problems.