Sep 14, 2026 · 25 min
AI turns insider risk into a machine-speed security problem
EP 73: A Teenager Could Now Run a Nation-State Attack — Insider Risk in the AI Era | Rajan Koo, DTEX
Organizations are adopting AI faster than they can govern employee behavior, shadow usage, and autonomous systems without resorting to intrusive surveillance.
- 1AI lets negligent or malicious insiders move faster and operate with far less technical expertise.
- 2Behavioral compliance, proportional monitoring, and privacy protections offer a more adaptive alternative to static policy enforcement.
- 3Layered defenses must combine data classification, access controls, user education, and continuous interpretation of AI behavior.
Don't miss
Rajan Koo explains why AI-driven attacks may operate at machine speed while defenders still depend on slower human workflows and constrained models.
The brief
Rajan Koo of DTEX frames the central shift: AI agents can accelerate insider-risk events so dramatically that attackers and careless employees may gain machine-speed advantages.
The episode argues that non-malicious misuse is becoming as important as deliberate attacks, because employees can expose sensitive data while simply trying to complete work with unapproved AI tools.
Koo’s alternative to blanket surveillance is behavioral compliance: collect only the data needed to quantify risk, protect privacy, and apply monitoring proportionate to demonstrated business exposure.
The sharpest security mismatch is speed: AI-driven attacks can outpace human analysts, while defenders may rely on weaker or restricted models to contain more capable systems.
Koo closes with a practical program: create a cross-functional AI steering committee, automate end-to-end processes, and layer guardrails around agents rather than trusting prompts alone.
Featuring
Mentioned
Hugging Face
Anthropic
MetaListen to the full episode and explore every guest, topic, and moment on PodLume.