Strategic HITL for Production AI Agents

A complete 13-part series on human oversight for AI agents, combining public research with hypothetical engineering designs and explicit limits on the evidence.

1. 00
   2026-07-01 Published

   Why Human-in-the-Loop is Not Optional for Production AI Agents

   The six-pillar argument for why HITL is not a transitional compromise but a structural requirement of any production AI system.

2. 01
   2026-09-22 Published

   AI and critical thinking: What the research actually shows

   What an EEG experiment, a knowledge-worker survey, and a perspective paper tell us about AI, critical thinking, and meaningful human oversight.

3. 02
   2026-09-23 Published

   The Hidden Debt in AI-Generated Code

   What public studies reveal about persistent code-quality issues, the limits of AI-versus-human comparisons, and reviewing architectural change.

4. 03
   2026-09-24 Published

   Graduated Automation: How to Scale AI Without Losing Control

   How a procedure earns permission to act: scoped approvals, independent checks, and withdrawing autonomy when conditions change.

5. 04
   2026-09-26 Published

   Your AI Agent's Skills Might Be Malware

   What the ToxicSkills audit found, what its numbers do not prove, and how to limit the damage from a compromised agent skill.

6. 05
   2026-09-26 Published

   When Human Oversight Pays for Itself

   A practical cost model for human review, with a hypothetical payback calculation, sensitivity checks, and cases where oversight does not break even.

7. 06
   2026-09-26 Published

   AI Governance Needs Shared Rules and Different Controls

   How to keep shared AI governance rules while tailoring controls to each use, with explicit ownership, testable evidence, and reassessment when permissions change.

8. 07
   2026-09-26 Published

   Five Ways to Question an AI Agent's Success

   A proposed review framework for checking an agent's diagnosis, scope, lasting outcome, downstream effects, and context freshness beyond a successful tool call.

9. 08
   2026-09-26 Published

   Designing Decision Boundaries for AI Autonomy

   How to turn autonomy policy into testable action conditions, using a hypothetical booking service to examine consent, reversibility, retries, and the cost of delay.

10. 09
    2026-09-26 Published

    What sleeper agents mean for production AI systems

    Sleeper-agent experiments show why cleaner evaluations do not prove harmful behaviour is gone, and why production checks need independent evidence.

11. 10
    2026-09-26 Published

    From vibe coding to verified: Practising independent diagnosis

    A proposal for practising independent diagnosis in AI-assisted engineering, with synthetic exercises and tests of whether the learning transfers.

12. 11
    2026-09-26 Published

    Incident resolution with AI: A hypothetical reference design

    A fictional incident workflow follows an AI-assisted rollback from alert to verified recovery, with explicit authority, handoff, and failure constraints.

13. 12
    2026-09-26 Published

    The road ahead: What evidence would justify more autonomy?

    Before removing an approval step, ask what evidence would justify the change, what the trial cannot show, and what would reverse the decision.