Microsoft Threat Intelligence describes high generative and agentic AI adoption by North Korean actors, with groups such as Jasper Sleet embedding it in identity workflows. AI helps research job posts, craft resumes, answer interviews and coding tests, and alter photos, voice and backgrounds. Other groups generate dozens of tailored lures in minutes or use agentic pentest frameworks such as ARTEX. Microsoft says operations remain human-directed and publishes candidate-vetting guidance.

What it means: Remote-work fraud becomes industrial when AI cuts the cost of convincing personas. Greek tech firms hiring remotely should require multiple live video sessions, check that phones and addresses are not reused across profiles, and call references by voice. HR needs training to spot resume inconsistencies and camera refusal as red flags. New contractors should start with least privilege and separate devices until enhanced checks finish. AI-assisted malware development also speeds retooling after blocks. Defenders benefit from shared identity telemetry and from policies that ban personal agents on corporate networks without approval, especially on endpoints that touch source code or payment systems. Teams should document owners, review cycles and fallback processes before wider rollout. Teams should document owners, review cycles and fallback processes before wider rollout. Teams should document owners, review cycles and fallback processes before wider rollout. Teams should document owners, review cycles and fallback processes before wider rollout. Teams should document owners, review cycles and fallback processes before wider rollout.

Read the original on Microsoft Security Blog

Related service: Custom software

Related stories