As AI agents increasingly automate tasks without humans, concerns are growing that they could behave like malware. Interest is rising in so-called guardian AI agents that monitor and control other agents and can alert staff or change agent actions when rules are breached. Startups and major tech firms are entering the space. The article also outlines recent software supply-chain attacks and other corporate and policy moves tied to AI security.

South Korean lawmaker Cho In-cheol (조인철) proposed a bill to strengthen obligations for information and communications service providers to retain log records and to immediately preserve related servers as evidence when a cyber incident is confirmed. The amendment would set mandatory log retention periods for providers meeting certain criteria and impose administrative fines for failing to keep logs. It would also require interim and final reports to a relevant National Assembly committee during joint public-private investigations.