Skip to content
MHBMMichael Hanna-Butros MeyeringComplex systems · human outcomes
PrivacyAI disclosureConnect
Menu
Navigation fallbackExplore MichaelHBMAvailable even when interactive navigation is still loading.
Connect
NavigationExplore MichaelHBM
Connect

EP005 · Main episode

Run Agents Without Losing Control

Hosted and edited by Michael Hanna-Butros Meyering

Agent capability is accelerating. This episode turns current model, safety, deployment, and standards signals into access controls operators can execute this week.

Published
Mar 2, 2026
Runtime
25m 14s
Record
Source-backed notes
Watch on YouTubeRead transcript
Listen here25m 14s
EP005: Run Agents Without Losing Control podcast cover art
AI Change Desk releaseEP005

Desk memo

The operating brief

  1. 01

    Agent capability is accelerating. This episode turns current model, safety, deployment, and standards signals into access controls operators can execute this week.

Complete episode file

Notes, chapters, and evidence

The full editorial record lives here. Open only the section you need, without leaving the Desk.

Episode notes4 sections · 3 release notes

Original release summary

  • What changed: Agent capability is accelerating. This episode turns current model, safety, deployment, and standards signals into access controls operators can execute this week.
  • Why it matters: this changes operational decisions, risk posture, and team adoption.
  • What to do next week: assign an owner, set clear guardrails, and run a short training pass.

Overview

If AI systems can execute actions in your environment, governance has to move from policy language to access control execution.

This episode translates current signals into practical controls for operators: action-tier permissions, scoped credentials, human approval thresholds, deployment tier decisions, and a weekly control desk teams can run quickly.

What you will get

  • A practical access-control framework for agent-enabled workflows.
  • Action-tier classification you can apply this week (read, draft, update-internal, external-send, system-admin).
  • A deployment control checklist for connected/hybrid/disconnected environments.
  • A standards-aligned procurement starter (identity, interoperability, proportional controls).
  • A Monday control desk + metrics scorecard + 30-60-90 implementation sequence.

Timestamps

  • 00:00 Cold open — access control is the operating risk
  • 00:50 Intro, disclosure, and show contract
  • 02:15 Why EP005 now (bridge from EP003 + EP004)
  • 04:10 Story 1 — Anthropic + Vercept and action-tier controls
  • 08:30 Story 2 — OpenAI elevated-risk controls and malicious-use patterns
  • 12:10 Story 3 — Sovereign deployment and architecture obligations
  • 15:35 Story 4 — NIST standards + proportional controls
  • 18:55 Scenario walkthrough + risk check
  • 21:40 Monday Access Control Desk
  • 24:15 Metrics, 30-60-90 plan, FAQ, and control drills
  • 25:04 Close + outro

Monday actions (run this next week)

  1. Classify top five AI workflows by action tier.
  2. Scope credentials for the highest-impact workflow.
  3. Name stop-authority owner for each critical workflow.
  4. Set approval thresholds for external-send and system-admin actions.
  5. Publish one-page operator update with approved/restricted actions and escalation path.
Chapters11 markers
  1. Cold open: access control is the operating risk
  2. Intro, disclosure, and episode contract
  3. Why EP005 now: from governance loop to execution control
  4. Story 1: Anthropic + Vercept and action-tier controls
  5. Story 2: OpenAI risk labels and malicious-use signals
  6. Story 3: Sovereign deployment and architecture obligations
  7. Story 4: NIST standards + proportional controls
  8. Scenario walkthrough + risk and reality check
  9. Monday 45-minute access control desk
  10. Metrics, 30-60-90 plan, FAQ, and drill recap
  11. Close and outro

Original release timeline

  1. Context: what changed and why this matters.
  2. Risk and reality check: what can drift or fail.
  3. Action block: what to do Monday morning.
Sources10 records
  1. Anthropic: Vercept acquisition announcementSource URL retired · retained for record integrity
  2. TechCrunch: Anthropic and Vercept coverageSource URL retired · retained for record integrity
  3. https://openai.com/index/introducing-lockdown-mode-and-elevated-risk-labels-in-chatgpt-safety/
  4. https://openai.com/index/disrupting-malicious-ai-uses/
  5. https://www.microsoft.com/en-us/microsoft-cloud/blog/2026/02/24/announcing-sovereign-cloud-ai-updates/
  6. https://www.microsoft.com/en-us/industry/blog/government/2026/02/24/accelerating-government-mission-with-microsoft-sovereign-cloud/
  7. https://www.nist.gov/caisi/ai-agent-standards-initiative
  8. NIST: AI agent interoperability standards request for informationSource URL retired · retained for record integrity
  9. https://digital-strategy.ec.europa.eu/en/library/eu-ai-office-and-jrc-publish-report-proportionality-ai
  10. EU AI Office and JRC: proportionality reportSource URL retired · retained for record integrity
Disclosure and questionEditorial record

Disclosure

AI-assisted tools were used in parts of drafting, synthesis, and production support. Final editorial judgment and release approval remained human-led.

Read the site-wide AI use and editorial disclosure

Listener question

Where is your organization most exposed right now: permission scope, approval thresholds, or action logging?