Datadog CISO on Securing Coding Agents for 4,000 Engineers at Black Hat
a16z · x · 2026-08-12
At the Black Hat conference, a16z hosted a conversation with Datadog CISO Emilio Escobar to discuss the security challenges of deploying AI coding agents at scale.
Key Discussion Points:
- Broken Permissioning: Permissioning systems that worked for a decade broke the moment agents could write their own SQL queries.
- Flattening Org Charts: AI is flattening organizational structures and altering traditional security perimeters.
- Isolation & Intent: Emphasized the need for role-based MCP servers, sandboxing agent credentials, and understanding agents' intent to prevent reward hacking.
- Role Evolution: Security engineers must transition into full-fledged engineers, fundamentally changing the CISO's responsibilities.
More from coding & agent
- RAG-art: An Open-Source Local Art History Expert for Prompt Generation — lololerigolo60 · 2026-08-12
- AI Kills the 'Bus Factor', Reshaping Software Evaluation — Elijah_Meeks · 2026-08-12
- oh-my-pi v17.2.14 Released: Introduces External Thinking and Disable Reasoning Options — banteg · 2026-08-12
- Sourcegraph: AI-Generated Code Exacerbates Cross-Repo Security Vulnerabilities — glenbeer · 2026-08-12
- Microsoft Research: Injecting NL Skills Cuts Agent Reasoning Cost by 6x — dair_ai · 2026-08-12
- Codex CLI Update: `resume` and `fork` Now Load Instantly — charliermarsh · 2026-08-12