CyberWire Daily Where encryption meets executive muscle.
9 snips
Dec 19, 2025 Nitay Milner, CEO of Orion Security, shares insights into data loss prevention in the AI era. He highlights the dangers of corporate data leaks into AI tools and discusses the importance of prioritizing DLP for CISOs. Milner addresses the limitations of legacy DLP methods, emphasizing the role of LLMs in improving detection accuracy and reducing false positives. He points out the dual nature of AI as a risk and a resource for enhanced data security, while anticipating new challenges in this rapidly evolving landscape.
AI Snips
Chapters
Transcript
Episode notes
Long-Nosed Goblin Exploited Group Policy
- ESET uncovered Long-Nosed Goblin using Group Policy to deploy malware and move laterally.
- Their Nosy Historian tool harvested browser data to find high-value victims in Southeast Asia and Japan.
LLMs Bring Nuance To DLP
- LLMs add human-like context to DLP, enabling nuanced verdicts beyond deterministic rules.
- This reduces false positives and makes DLP operate more like EDR for data, per Nitay Milner.
DLP Needs An EDR-Like Evolution
- Traditional rule-based DLP and UEBA produced many false positives because human behavior is variable.
- Treating data protection like endpoint detection and response unlocks better scalability and efficacy.