EmergingAI control, monitoring and containment

AI Shutdown and Recovery Engineer

Also called Agent Intervention Engineer

An AI Shutdown and Recovery Engineer builds and tests the ways to stop an AI system mid-action and bring it back to a known good state. It matters because untrusted model behaviour must never turn into damage elsewhere.

What they do

They design interruption paths for agents and other bounded systems, then test whether those paths actually work. That means deciding what can be stopped, by whom, and how fast. They build kill switches, set up sandboxes and write monitors that trigger intervention. They work out what state a system must return to, and how to restore it without repeating or half-finishing actions. They run recovery drills, attack their own controls with adversarial evaluation harnesses, and write procedures that an on-call engineer can follow under pressure. They fix the gaps the drills expose.

What a strong hire delivers

In the first months, a strong hire maps where the company's agents can take action and which of those actions can be interrupted. They deliver a working intervention path for the highest-risk system, written interruption procedures, and a first recovery drill with honest results. They leave behind a repeatable test that shows whether intervention still works after each change.

How you can tell they are good

Ask them to walk through an intervention they built and how they proved it worked. Good candidates talk about failure first: partial actions, race conditions, a system that resists being stopped. Ask what they would do if a monitor fired during a multi-step task. Be wary of anyone who only describes a stop button.

Why it is hard to fill

Almost nobody holds this title yet. The work sits inside security, safety and platform roles, so CVs rarely say it. Candidates usually have deep strength in one area, such as security or formal methods, and thin exposure to model behaviour. Recruiters have to read for the work done, not the job title, and judge how well someone bridges the gaps.

Where to find them

Look in security engineering teams with incident response and containment experience. Look at people from formal methods and verification, and from AI safety evaluation or red teaming groups. Frontier labs, safety-focused start-ups and companies running autonomous agents in production are the likeliest sources. Some come from reliability and infrastructure work.

Indicative pay

Not enough published evidence yet to give a fair range for this role, which is common for work this new. We benchmark it for you as part of a search.

Interview questions for an AI Shutdown and Recovery Engineer

  1. Tell us about a system you built to stop or contain something that might misbehave. How did you test it?

    Strong answers describe a real mechanism, a clear threat model and tests that tried to defeat it. Worry if testing was only a happy-path demo or if they cannot say what the mechanism failed to cover.

  2. An agent is halfway through a multi-step task when you interrupt it. What state is the system in, and how do you recover?

    Look for talk of partial actions, idempotency, checkpoints, rollback and compensating steps. They should ask what the agent could touch. Worry if they assume stopping is clean and instant.

4 more questions, with what a strong answer covers, are in the Role Kit.

Job description you can post

We are hiring an engineer to make sure we can stop our AI systems and recover them safely. You will design intervention paths, test them against adversarial behaviour and run the drills that prove they work. The work is new, so we value judgement and evidence of building things that held up.

What you will do

  • Design and build interruption paths and kill switches for agents and bounded systems
  • Define safe recovery states and build the tooling to restore them
  • Run regular recovery drills and report results honestly
  • Build adversarial evaluation harnesses that try to defeat our controls
  • Set up sandboxes and monitors that trigger intervention
  • Write clear procedures for on-call engineers
  • Work with research and product teams to fix gaps found in testing

What you will bring

  • Shipped intervention, containment or incident response mechanisms that were tested under pressure
  • Experience in security engineering, formal methods or AI safety evaluation
  • Ability to reason about partial failure, state and rollback
  • Hands-on work with sandboxing and monitoring
  • Experience testing systems adversarially
  • Clear written procedures that others have used

Nice to have

  • Formal verification of system properties
  • Experience with agent frameworks in production
  • Background in AI control research

Free to use and adapt. If it saves you time, a link back to this page is appreciated.

Skills that matter

  • AI controlAdvancedLimit harm from potentially untrusted AI behaviour
  • Security engineeringAdvancedProtect systems, identities and sensitive assets
  • Model evaluationWorkingMeasure capability and failure modes credibly
  • Formal verificationWorkingProve stated properties under explicit assumptions

Other roles in aI control, monitoring and containment

EmergingAI Control ResearcherAn AI Control Researcher designs the safeguards that limit what a model can do if it turns out to be unreliable or hostile. They keep risk bounded…EmergingAgent Containment EngineerAn Agent Containment Engineer builds the boundaries that keep AI agents from doing harm, even if the model misbehaves or is attacked. They matter…EmergingAI Monitor Evaluation EngineerAn AI Monitor Evaluation Engineer tests whether the systems that watch model behaviour actually catch dangerous or unauthorised actions. It matters…EmergingAI Control EngineerAn AI Control Engineer builds the monitoring, sandboxing and containment that let a company use powerful models safely, even if a model is…

Related roles in other families

EmergingAI Formal Verification EngineerFormal methods and reliable reasoningAn AI Formal Verification Engineer writes precise specifications for AI systems and their outputs, then proves they hold or finds counterexamples…Hiring nowAutomated Theorem Proving ResearcherFormal methods and reliable reasoningAn Automated Theorem Proving Researcher builds and improves methods that find or check mathematical and logical proofs by machine. The company gets…Hiring nowProgram Synthesis ResearcherFormal methods and reliable reasoningA Program Synthesis Researcher builds systems that generate programmes which meet a testable or formal specification. They give the company code…EmergingAI-Assisted Verification SpecialistFormal methods and reliable reasoningAn AI-Assisted Verification Specialist uses AI tools to speed up proof and specification work, then checks every result independently. They help a…

How we wrote this

Role definitions are our own editorial work, checked against occupational classifications, standards, research and employer pages. Titles vary by company, and this is a description of the work, not a job advert.

← AI control, monitoring and containmentAll new job families

Start a search

Hiring an AI Shutdown and Recovery Engineer?

Tell us the work. We find the people already doing it, whatever their job title says, and you meet them already interviewed for your role.