Skip to main content

Authority Path — Concept Review

Date: 2026-02-27 Trigger: Notion "Authority Paths Primer" — synced 2026-02-27 Status: Strategic review with critical feasibility analysis


The Primer's Core Thesis

From authority-paths-primer.md (Notion, synced today):

We must model authority paths as execution-determined, not configuration-determined.

Three distinct layers:

LayerWhat it capturesExample
1. PotentialAll possible permissions/roles that could executeSP has Contributor + Reader roles on 5 resources
2. Actual executionWhat did execute — APIs called, tokens used, scopes invoked, resources touchedSP used Reader role to call GET /users on Microsoft Graph
3. Standing authority / risk postureThe delta between 1 and 2SP has Contributor but only ever used Reader — latent write authority exists

Default display rule: Only show what was exercised. Latent authority is a separate "could also do" view.

This creates two abstractions:

  • Observed Authority Path — what actually executed (default view)
  • Potential Authority Path — what could execute (future addition)

The differentiation lives in (2) — actual execution with provenance and execution lineage.


1. SEO Perspective — Strengthened by Execution Focus

The execution-determined framing sharpens the competitive differentiation:

CompetitorTheir framingSecurityV0 after Primer
SpecterOps/BloodHound"What an attacker COULD exploit" (potential)"What autonomous systems DID execute" (observed)
Veza"Who CAN access what" (potential)"What DID access what, under what authority" (observed)
Wiz"What toxic combinations EXIST" (potential)"Which toxic combinations EXECUTED" (observed)
CrowdStrike"What threats are DETECTED" (reactive)"What authority was EXERCISED" (proactive)

Key SEO angle: Every competitor shows potential. SecurityV0 shows actual. This is the "did access" vs "could access" positioning from the Clarity doc.

New content opportunities:

  1. "Observed Authority vs Potential Authority: Why Most Security Tools Show the Wrong Thing"
  2. "Execution-Determined Security: Moving Beyond Configuration Scanning"
  3. "The 90% Noise Problem: Why Potential Authority Paths Are Misleading"

Category-creation angle: "Execution-Determined Exposure Management" — nobody owns this.


2. Platform Perspective — Concept Is Sound, Execution Evidence Is the Bottleneck

The Primer aligns well with recent product thinking

The Clarity doc (clarity-so-what-what-do-i-do-now.md) reinforces: "Everything in the CISO view must be execution-determined, not configuration-determined."

The Sprint doc confirms this is the current focus: finish W1 integration + add clarity narratives.

What changes from the current architecture

Current (glossary)After Primer
Authority Path = 4-node projection (Workload → Identity → Destination → Data Domain)Authority Path splits into Observed (execution-derived) and Potential (configuration-derived)
Computed from execution_paths[] (which are role → permission → resource chains)Observed path must be computed from execution evidence, not just role assignments
Single viewTwo views: exercised branch only (default) vs full potential (secondary)

The 4-node model survives but needs qualification

The Workload → Identity → Destination → Data Domain projection still works, but now:

  • Observed path: Only includes the identity/role/destination that was actually used in execution
  • Potential path: Shows all possible routes through all assigned roles

Example: If a SP has roles Contributor and Reader but only exercised Reader to reach Microsoft Graph:

Observed:  Workload → SP (via Reader) → Microsoft Graph → Identity Platform
Potential: Workload → SP (via Reader + Contributor) → Microsoft Graph + Key Vault + Storage

3. Customer / CSR Perspective — Execution Focus Resonates Strongly

The Clarity doc captures the customer voice perfectly:

"I understand what you're doing, but you're making me think too much."

The execution-determined framing solves this by eliminating noise:

ProblemHow execution-focus fixes it
Too many paths shown (potential authority creates noise)Show only paths that actually executed — 90% noise reduction
"So what?" — configuration doesn't prove riskExecution evidence proves risk is real, not theoretical
CISOs don't trust "could" analysis"This DID happen 14,000 times this week" is undeniable

The risk narrative from the Clarity doc maps perfectly:

  1. What happened — Standing autonomous authority exists + X executions + sensitive domain touched
  2. Why it matters — Orphaned owner + scope drift + LLM egress
  3. What reduces risk — Assign owner, reduce RBAC scope, restrict endpoint

Market positioning upgrade: "Most NHI tools show you what your identities COULD do. SecurityV0 shows you what they DID do — with deterministic proof."


4. Documentation Perspective — Term Validation Updated

The Primer doesn't change the term "Authority Path" — it sharpens what it means. The key documentation update needed:

Current glossary definition:

Authority Path: A 4-node linear projection: Workload → Identity → Destination → Data Domain

Proposed update:

Authority Path: A 4-node linear projection showing execution authority from Workload → Identity → Destination → Data Domain. Two forms: Observed (derived from execution evidence — what actually executed) and Potential (derived from configuration — what could execute). The default view is Observed. The structural backbone of an Exposure.


5. Competitive Positioning (Updated One-Liners)

AgainstUpdated positioning
SpecterOps"They simulate what attackers COULD do. We show what autonomous systems DID do — with provenance."
Veza"They map who CAN access what. We map what DID access what, how often, and under whose authority."
Wiz"They find toxic combinations that COULD be exploited. We find authority that WAS exercised — without an owner."
All competitors"Every tool shows potential authority. We show observed authority. That eliminates 90% of the noise."

6. Action Items (Updated)

PriorityAction
P0Update glossary: split Authority Path into Observed vs Potential
P0Validate execution evidence feasibility (see companion review)
P1Update path detail UX: default to observed (exercised branches only)
P1Define "Potential Authority" as secondary view with clear "could also do" framing
P2Create public concept page: "Observed vs Potential Authority Paths"
P2Map to OWASP NHI Top 10 with execution-determined framing

Sources

  • Authority Paths Primer (Notion, synced 2026-02-27)
  • Clarity: So What? / What Do I Do Now? (Notion, synced 2026-02-27)
  • Sprint: 2026-02-25 – 2026-03-12 (Notion, synced 2026-02-27)
  • [DRAFT] Drift Intelligence W1 (Notion, synced 2026-02-27)
  • [DRAFT] Actionability / Remediation Guidance W1 (Notion, synced 2026-02-27)
  • SEO/competitive research (web, 2026-02-27)
  • Connector execution evidence analysis (codebase, 2026-02-27)