Getting AI agents past enterprise security reviews is harder than building them
Useful_Lecture_5927 · reddit · 2026-09-01
A developer building agents for enterprise environments says the hardest part isn't making the agent work — it's getting security teams comfortable letting it touch CRM data, email, internal APIs, databases, and MCP servers.
The recurring questions: what exactly can the agent access, what happens under prompt injection, and can you audit what it did after something goes wrong. He asks practitioners for ugly lessons: middleware/control layers that actually block tool calls, what audit trails look like, identity and least-privilege across agents/users/tenants, and the biggest gaps reviewers found after teams thought they were covered.
More from coding & agent
- Fintech HR built a sourcing agent on 60,000 past applicants — 30 qualified candidates in a month — G0NZ0BOY · 2026-10-16
- 3x DGX Spark cluster runs GLM-5.3-Flash locally behind a proxy, 1.75M tokens/day at 119W — Teknium · 2026-10-16
- Builder's one rule for trusted desktop agents: read freely, but sending, deleting, buying always ask first — No_Fan5310 · 2026-10-16
- GPT-built unifi-simulator makes MikroTik switches appear inside UniFi Network — Pytlicek · 2026-10-16
- Stanford drops 45-min SpaceX lecture on how Grok Bot's always-on agent harness works — Hesamation · 2026-10-16
- Opus 5.5 + Three.js demo stuns with water physics, fish and blossoms — nikola_mr64990 · 2026-10-16