A practical guide to monitoring AI agents in production: the four signal groups that matter, run tracing, quality drift detection, prompt-injection and tool-abuse signals, cost per successful outcome, and a staged 90-day rollout.
AI system inventory work starts with an uncomfortable question nobody can answer: which AI tools are we actually using? This guide gives you the register that answers it — a field-by-field template, the five places shadow AI genuinely hides, seven discovery methods that surface tools nobody declared, a two-week sweep plan, a four-tier triage scale, and a clean mapping to ISO/IEC 42001, the NIST AI RMF and the EU AI Act.
AI risk assessment is the step most organisations skip between “this tool looks useful” and “forty people now depend on it”. This guide gives you a two-page template you can fill in section by section, the seven risk categories worth scoring, a likelihood-impact-exposure scale that produces comparable numbers, the six steps of the exercise itself, the five controls that genuinely move a score, and a clean mapping to the NIST AI RMF, ISO/IEC 42001 and the EU AI Act.
AI vendor lock-in is the bill that arrives eighteen months after a successful pilot. This guide breaks dependency into five distinct types — model, data, workflow, integration and commercial — explains why AI switching costs behave differently from classic software migrations, sets out the contract clauses that cap your exposure at signature, covers the architecture decisions that keep switching cheap, and gives you a four-page exit plan you can write in a day and rehearse once a year.
Your staff are already using AI at work; the only question is whether they are doing it inside rules you wrote. This guide gives you a complete AI acceptable use policy template for employees: the nine clauses that actually carry weight, model wording you can copy for scope, approved tools, data entry, human accountability, disclosure and breach handling, a three-tier approved and prohibited tool model, a data table showing what may never be pasted into a public chatbot, a rollout plan that gets the policy acknowledged, the mistakes that quietly kill enforcement, and the four numbers that tell you whether any of it is working.
ISO 42001 certification costs a UK organisation roughly £12,000 to £180,000 in year one and takes six to fifteen months, and neither range means anything until you know what moves it. This guide splits the cost into the four budgets hiding behind one number, shows how certification bodies calculate audit days under ISO/IEC 42006, sets out a month-by-month timeline from gap analysis to certificate decision, explains the five things that reliably push the date, models the three-year cost that matters more than year one, and lists six levers that cut spend and elapsed time without weakening the certificate.
Once a year is a floor, not a schedule. This guide sets out how often a business should conduct penetration testing and why the calendar date matters far less than what changed in the estate since the last report. It covers the twelve-month baseline and where it comes from, the seven change triggers that should force an unscheduled round, exactly what PCI DSS, ISO 27001, SOC 2, Cyber Essentials Plus and NIS2 actually require, where vulnerability scanning stops and human testing starts, indicative UK programme costs at every cadence, and how to build a calendar that survives a year of competing priorities.
Most cybersecurity risk register templates are built for banks and abandoned by small businesses within a fortnight. This guide strips the document back to the eleven fields that earn their place, gives likelihood and impact scales anchored to time and money rather than adjectives, and shows a worked register for a sixty-person firm with real rows, owners and treatment decisions. It also covers the four treatment options and how to use each one honestly, a two-afternoon build method, the review cadence and out-of-cycle triggers that stop the register rotting, and when a spreadsheet stops being enough.
An incident response retainer is a contract you buy before anything has happened, to guarantee access to specialists who are otherwise fully booked the moment a large ransomware event hits the market. The cheapest and most expensive quotes can describe genuinely different products, and on a procurement spreadsheet they look interchangeable. This guide covers what you are actually buying: the standard reactive and proactive inclusions, the exclusions that destroy budgets, the three pricing models in common use, realistic UK cost bands for 2026 by organisation size, what response-time service levels genuinely promise, how prepaid hours are consumed and lost, and a scorecard for comparing providers before you sign.
An incident response plan that has never been tested is a document, not a capability. A cyber tabletop exercise is the cheapest way to find out whether your organisation can actually respond — who holds shutdown authority, when the regulatory clock starts, and whether anyone has drafted a holding statement before they needed one. This guide covers the full cycle: setting objectives and scope, choosing a scenario grounded in your real risk register, deciding who belongs in the room, building the four-document exercise pack, a three-hour run sheet, the facilitation techniques that keep the discussion honest, and the after-action reporting that converts findings into tracked and closed actions.
