--- title: "A Control Blueprint for AI: 2026 Playbook | Firevault" description: "A 40-page 2026 playbook on AI infrastructure, open weights, agents, kill-switch design and operational recovery for boards, CISOs and AI leaders." lang: en-GB json-ld: | [ { "@context": "https://schema.org", "@type": "Organization", "@id": "https://fire-vault.com/#organization", "name": "Firevault", "legalName": "Firevault Limited", "url": "https://fire-vault.com", "logo": { "@type": "ImageObject", "url": "https://fire-vault.com/logo.png", "width": 200, "height": 60 }, "foundingDate": "2025-03", "description": "Firevault has rewritten the rules for data protection & storage via its offline secure storage (OSS) platform, for users to vault everything that matters.", "address": { "@type": "PostalAddress", "addressCountry": "GB", "addressLocality": "United Kingdom" }, "contactPoint": [ { "@type": "ContactPoint", "contactType": "customer service", "email": "hello@fire-vault.com", "availableLanguage": "English", "areaServed": "GB" } ], "sameAs": [ "https://www.linkedin.com/company/firevault", "https://x.com/firevaultuk" ], "slogan": "Disconnect to Protect", "knowsAbout": [ "Offline Secure Storage", "Physical Air Gap Data Protection", "Ransomware Protection", "Data Sovereignty", "GDPR Compliance", "NIS2 Compliance" ] }, { "@context": "https://schema.org", "@type": "WebSite", "@id": "https://fire-vault.com/#website", "name": "Firevault", "alternateName": [ "Firevault", "Firevault UK", "Firevault Limited" ], "url": "https://fire-vault.com", "publisher": { "@id": "https://fire-vault.com/#organization" }, "inLanguage": "en-GB", "description": "Firevault has rewritten the rules for data protection & storage via its offline secure storage (OSS) platform, for users to vault everything that matters.", "potentialAction": { "@type": "SearchAction", "target": { "@type": "EntryPoint", "urlTemplate": "https://fire-vault.com/learn?q={search_term_string}" }, "query-input": "required name=search_term_string" } }, { "@context": "https://schema.org", "@type": "WebPage", "@id": "https://fire-vault.com/playbook/ai-control-blueprints#webpage", "url": "https://fire-vault.com/playbook/ai-control-blueprints", "name": "A Control Blueprint for AI: 2026 Playbook", "description": "A 40-page 2026 playbook on AI infrastructure, open weights, agents, kill-switch design and operational recovery for boards, CISOs and AI leaders.", "isPartOf": { "@id": "https://fire-vault.com/#website" }, "about": { "@id": "https://fire-vault.com/#organization" }, "primaryImageOfPage": { "@type": "ImageObject", "url": "/__l5e/assets-v1/d10e0e16-e107-4d93-b690-87f7b33bea9b/ai-control-cover.jpg" }, "inLanguage": "en-GB", "breadcrumb": { "@id": "https://fire-vault.com/playbook/ai-control-blueprints#breadcrumb" } }, { "@context": "https://schema.org", "@type": "BreadcrumbList", "@id": "https://fire-vault.com/playbook/ai-control-blueprints#breadcrumb", "itemListElement": [ { "@type": "ListItem", "position": 1, "name": "Home", "item": "https://fire-vault.com" }, { "@type": "ListItem", "position": 2, "name": "Home", "item": "https://fire-vault.com" }, { "@type": "ListItem", "position": 3, "name": "Playbooks", "item": "https://fire-vault.com/playbooks" }, { "@type": "ListItem", "position": 4, "name": "A Control Blueprint for AI: 2026 Playbook", "item": "https://fire-vault.com/playbook/ai-control-blueprints" } ] }, { "@context": "https://schema.org", "@type": "Article", "headline": "A Control Blueprint for AI: 2026 Playbook", "description": "A 40-page 2026 playbook on AI infrastructure, open weights, agents, kill-switch design and operational recovery for boards, CISOs and AI leaders.", "inLanguage": "en-GB", "author": [ { "@type": "Person", "name": "Mark Fermor" } ], "publisher": { "@type": "Organization", "name": "Firevault", "url": "https://fire-vault.com" }, "image": "/__l5e/assets-v1/d10e0e16-e107-4d93-b690-87f7b33bea9b/ai-control-cover.jpg", "keywords": "AI control blueprint, AI kill switch, model weight security, AI agent governance, AI infrastructure control plane, responsible AI operations", "mainEntityOfPage": "https://fire-vault.com/playbook/ai-control-blueprints" }, { "@context": "https://schema.org", "@type": "FAQPage", "mainEntity": [ { "@type": "Question", "name": "What makes this a 'control blueprint' rather than an AI policy?", "acceptedAnswer": { "@type": "Answer", "text": "Policies describe intent. A control blueprint describes decisions: where AI can run, what it can reach, which assets it can change, who can stop it, and how you recover if the system fails. The playbook is structured around those decisions and the evidence needed to prove them." } }, { "@type": "Question", "name": "Who should read A Control Blueprint for AI?", "acceptedAnswer": { "@type": "Answer", "text": "Boards, audit committees, CISOs, CIOs, chief AI officers, AI programme leads and infrastructure teams. It is written to be read by both the leadership and the operators who will implement the controls." } }, { "@type": "Question", "name": "What is the AI kill switch doctrine?", "acceptedAnswer": { "@type": "Answer", "text": "It is a design pattern for controlled stop states, from throttling an agent, through model withdrawal, to a full hard stop, where the stop path is independent of the system being stopped. Part four of the playbook covers it in detail." } }, { "@type": "Question", "name": "Which research does the playbook draw on?", "acceptedAnswer": { "@type": "Answer", "text": "AISI's Frontier AI Trends Report, NCSC frontier AI guidance, NIST AI RMF, OWASP GenAI and agentic guidance, MITRE ATLAS, RAND's Securing AI Model Weights work and the International AI Safety Report 2026." } }, { "@type": "Question", "name": "Does the playbook recommend specific tooling?", "acceptedAnswer": { "@type": "Answer", "text": "No. It is deliberately vendor-neutral. It sets out the control decisions and evidence a mature AI operating model needs, and leaves tooling choices to your architecture team." } }, { "@type": "Question", "name": "How long is it and how is it delivered?", "acceptedAnswer": { "@type": "Answer", "text": "40 pages, roughly 28 minutes to read. Request it with your work email and a secure link arrives within a minute; access is tied to that address." } } ] }, { "@context": "https://schema.org", "@type": "BreadcrumbList", "itemListElement": [ { "@type": "ListItem", "position": 1, "name": "Home", "item": "https://fire-vault.com" }, { "@type": "ListItem", "position": 2, "name": "Playbooks", "item": "https://fire-vault.com/playbooks" }, { "@type": "ListItem", "position": 3, "name": "A Control Blueprint for AI: 2026 Playbook", "item": "https://fire-vault.com/playbook/ai-control-blueprints" } ] } ] --- Recent Breaches Breaches [2026 PowerSchool 62.4M records ](https://www.bleepingcomputer.com)[2026 DISA Global Solutions 3.3M records ](https://techcrunch.com)[2026 Globe Life 850K records ](https://www.securityweek.com)[2026 Lidl GB Customer contact data ](https://www.theguardian.com/business/2026/jun/11/lidl-gb-third-party-data-breach)[2026 Asahi Group Production systems disrupted ](https://www.reuters.com/technology/cybersecurity/asahi-says-cyberattack-disrupted-operations-japan-2025-09-29/)[2026 Kido International 8K records ](https://www.bbc.co.uk/news/articles/c623d7v0e5xo)[2026 Collins Aerospace (RTX) Check-in and boarding disruptio... ](https://www.bbc.co.uk/news/articles/c789e7l1z7po)[2026 Jaguar Land Rover Production and IT systems disru... ](https://www.bbc.co.uk/news/articles/cx2gx8p3rzeo)[2026 Peter Green Chilled Order and logistics data ](https://www.bbc.co.uk/news/articles/c0k7yy8n3g5o)[2026 Adidas UK Customer contact details ](https://www.bbc.co.uk/news/articles/c78jkev1el2o)[2026 PowerSchool 62.4M records ](https://www.bleepingcomputer.com)[2026 DISA Global Solutions 3.3M records ](https://techcrunch.com)[2026 Globe Life 850K records ](https://www.securityweek.com)[2026 Lidl GB Customer contact data ](https://www.theguardian.com/business/2026/jun/11/lidl-gb-third-party-data-breach)[2026 Asahi Group Production systems disrupted ](https://www.reuters.com/technology/cybersecurity/asahi-says-cyberattack-disrupted-operations-japan-2025-09-29/)[2026 Kido International 8K records ](https://www.bbc.co.uk/news/articles/c623d7v0e5xo)[2026 Collins Aerospace (RTX) Check-in and boarding disruptio... ](https://www.bbc.co.uk/news/articles/c789e7l1z7po)[2026 Jaguar Land Rover Production and IT systems disru... ](https://www.bbc.co.uk/news/articles/cx2gx8p3rzeo)[2026 Peter Green Chilled Order and logistics data ](https://www.bbc.co.uk/news/articles/c0k7yy8n3g5o)[2026 Adidas UK Customer contact details ](https://www.bbc.co.uk/news/articles/c78jkev1el2o) [View All →](/learn/breaches) [![Firevault - offline secure storage, physically disconnected from the internet](/assets/logo-color-DBVl0KCg.png)](/) Products Solutions [Why OSS](/why-oss) More [Help](/help)[Get started](/get-started) 1. [Home](/) 2. [Playbooks](/playbooks) 3. A Control Blueprint for AI A Control Blueprint for AI # A Control Blueprint for AI Control the path. Govern the agent. Protect the asset. A 40-page 2026 playbook on AI infrastructure, open weights, agents, kill-switch design and operational recovery for boards, CISOs and AI leaders. 40 pages 28 min read Boards and audit committees ![A Control Blueprint for AI: 2026 Playbook cover](/__l5e/assets-v1/d10e0e16-e107-4d93-b690-87f7b33bea9b/ai-control-cover.jpg) ## Get the playbook Emailed to you within a minute First name \* Last name \* Work email \* Personal email addresses such as Gmail or Yahoo are not accepted. Company \* Sector \*Select one, ... Job title \* I agree to receive the playbook and occasional Firevault briefings. See our [privacy policy](/privacy-charter). Email me and open the playbook Email-gated GDPR compliant 01 Foreword ## Written by the people who build the controls. AI governance is moving quickly, and the danger is treating it as a paperwork exercise while agents, models and infrastructure become part of the operating fabric of the business. A Control Blueprint for AI is a deliberately different document. It is not a model-security checklist; it is a control blueprint for the people who now have to decide where AI runs, what it can reach, which assets it can change, who can stop it, and how the business recovers if the system fails or is abused. Grounded in AISI, NCSC, NIST AI RMF, OWASP GenAI, MITRE ATLAS, RAND and the International AI Safety Report 2026, and written by Mark Fermor, it gives boards and CISOs a common vocabulary, Control the Path, Govern the Agent, Protect the Asset, and the operating cadence to run AI control as a discipline, not a policy. Every Firevault playbook is grounded in the same doctrine behind our Offline Secure Storage® platform: control the path, govern the agent, protect the asset. **Mark Fermor** · Co-Founder, Firevault 02 What's inside ## 6 parts. Written for the people who own the decision. Each chapter opens with the decision it exists to help you make, and closes with the evidence you should expect back. 01 ### The AI infrastructure reality Why AI is now critical infrastructure, and how control planes, data planes and recovery planes must be separated across compute, data, identity, facilities and recovery. - AI infrastructure is now a national security issue - One shared path means one shared failure 02 ### Attack surface and the AI data centre Chained attack surface analysis, prompt, identity, tool, API, data, recovery, and the concentration risk sitting inside modern AI clusters. - Attackers rarely need to own the model if they own the route - Model theft is now a data-centre security problem 03 ### Open weights and model assets Protecting weights like source code plus crown jewels: custody, safeguards, release control, hashing, offline archives and destruction records. - Open-weight releases change the risk boundary permanently - Model custody is a lifecycle discipline, not a one-off decision 04 ### Agents and action authority Identity, scope, purpose, expiry, approval gates and revocation for AI agents. Why recommendation is not delegated authority. - Agents need identity and expiry, like people do - Control the tool, not only the prompt 05 ### The AI kill switch doctrine Controlled stop states, agent stop, model withdrawal and the hard stop. Designing the safe landing zone and proving it works. - A kill switch must have levels, not just an off button - The deeper the consequence, the more independent the stop path must be 06 ### Policy, adoption and operating cadence From principles to operations. Five board questions, control outputs, a pilot pattern with a control boundary, and a working operating cadence for AI control. - A pilot is judged on reach, authority, evidence and recovery - AI control is an operating rhythm, not a policy binder 03 Who it's for ## Read it if you're accountable for the decision. - Boards and audit committees - CISOs, CIOs and chief AI officers - AI programme and platform leads - Data-centre and infrastructure teams Sectors we hear from Financial services Government & defence Critical national infrastructure Technology & AI Healthcare & life sciences 05 Frequently asked ## Questions leaders ask before requesting. Something else on your mind? Reply to any Firevault email or write to [mark@fire-vault.com](mailto:mark@fire-vault.com). ### What makes this a 'control blueprint' rather than an AI policy? ### Who should read A Control Blueprint for AI? ### What is the AI kill switch doctrine? ### Which research does the playbook draw on? ### Does the playbook recommend specific tooling? ### How long is it and how is it delivered? 06 More Firevault playbooks ## Board-level control blueprints. Each playbook is written for the people who own the decision, practical, UK-grounded and free to request. [ ![The Leaders' Playbook cover](/__l5e/assets-v1/e8c2b388-0083-4aa9-a143-f8c4b11d2db3/leaders-cover.jpg)](/playbook/leaders) Related playbook ### The Leaders' Playbook A board-level briefing on sovereign data, succession resilience and physical protection. 28 pages 18 min read Email-gated [Get the playbook](/playbook/leaders) [ ![Firebreak for Water: Adoption & Deployment Playbook cover](/__l5e/assets-v1/87198e25-711c-49ab-95d1-5f7cc99553cb/firebreak-water-cover.jpg)](/playbook/firebreak-water) Related playbook ### Firebreak for Water: Adoption & Deployment Playbook Layer 1 physical path control mapped against the Purdue model for water operators. 28 pages 22 min read Email-gated [Get the playbook](/playbook/firebreak-water) [ ![Close the File. Protect the Record. cover](/__l5e/assets-v1/a563c7e7-cdd3-47c3-8557-57a4d04b97b7/legal-cover.jpg)](/playbook/legal) Related playbook ### Close the File. Protect the Record. File closure, retained records and offline custody for UK law firms. 28 pages 20 min read Email-gated [Get the playbook](/playbook/legal) [ ![A Control Blueprint for Aerospace & Aviation cover](/__l5e/assets-v1/bb5110b1-edcd-4a3a-93fc-af134ae6e4b7/aerospace-cover.jpg)](/playbook/aerospace) Related playbook ### A Control Blueprint for Aerospace & Aviation Communication registers, operational states and rapid isolation for aerospace and aviation. 28 pages 22 min read Email-gated [Get the playbook](/playbook/aerospace) ## Ready to read it? Request A Control Blueprint for AI. It lands in your inbox within a minute, tied to your email address. No download, no fuss. [Get the playbook](#get-the-playbook)[Talk to Mark](mailto:mark@fire-vault.com)