SANS 2026 Identity Threats Survey
AssessTools
A security survey cited for data on non-human identities and AI credential usage.
Why it's here
Placed in Assess: 1 article(s) of evidence from 1 source(s), led by security coverage, with 1 in the last 30 days. Confidence 24%. Low accumulated evidence, so it defaults conservatively pending more signal.
Evidence (1)
- 8The New Stack·9/12/2026securityMCP security shifts toward permission controls
The article argues that the main security issue with Model Context Protocol (MCP) is not the protocol itself, but overly broad permissions and credentials granted to AI agents and integrations. It cites recent incidents and survey data to show that teams need tighter scoping, temporary credentials, and action-based authorization rather than trusting default access.