TrustedIntelligence
The AI Governance Playbook for Growing Regulated Businesses
Your organisation is already using AI. Some of it you sanctioned. Some of it you did not. All of it carries risk.
This is the practical, jurisdiction-aware guide to governing the AI you actually have. Not the AI that appears in conference presentations.
No spam. Early access, launch discounts, and governance resources.
Your AI governance programme has not caught up with your AI use.
The mid-market enterprise in 2026 faces five overlapping AI realities. None of them are waiting politely for a governance framework to arrive.
Too large to be ignored by regulators. Too lean to have a dedicated AI governance team. Too commercially pressured to slow down. This book was written precisely for that position.
Generative AI is everywhere
Every employee with access to Microsoft 365 Copilot, ChatGPT, or Gemini is using a system that produces non-deterministic outputs. Your organisation has not assessed most of them.
Copilots blur accountability
When a copilot drafts advice that an employee sends to a client, who is accountable for its accuracy. Your current governance framework probably does not say.
RAG creates new data risks
Retrieval-augmented generation puts new pressure on what data you hold, what enters the retrieval index, and what comes back in responses. Most organisations have not mapped that exposure.
Agents take real-world action
AI that plans multi-step tasks and executes them without human intervention at each step is entering the enterprise. Defining what an agent can do, and ensuring it does not exceed that boundary, is a governance challenge most organisations are not yet equipped to handle.
Shadow AI is your biggest unmanaged risk
Employees across every function are using AI tools the organisation has not sanctioned, assessed, or even identified. It is not malicious. It is the natural response of capable people to powerful free tools. But it is carrying risk your governance framework is not covering.
Written for the people who have to make the decision.
Not data scientists. Not AI engineers. Senior leaders and governance professionals at mid-market enterprises in regulated sectors, who need to act and need to be able to justify that action to a board.
Senior Leaders
CEOs, CROs, CCOs, CFOs, and Heads of Digital Transformation at organisations with 250 to 5,000 employees and £50m to £500m revenue in regulated sectors. Intelligent, time-pressured, commercially minded, and done with vague governance advice.
Governance Professionals
AI governance consultants, risk professionals, and compliance officers who work with mid-market clients and need a credible, structured framework they can implement alongside their clients. Not just present to them.
Board Members
Non-executive directors and board members who need enough grounding to ask the right questions, hold management accountable, and understand whether the assurances they are receiving are credible.
Finish this book on a Friday.
Brief your board on Monday.
Every chapter leaves you with something concrete. Not theory to sit with. Tools to use, frameworks to apply, and decisions you can make before the week is out.
A complete picture of your AI exposure
The AI Exposure Diagnostic shows you exactly which categories of AI risk your organisation carries. Including the AI you did not sanction and the vendors who shipped it to you anyway.
Clarity on what regulations actually require
EU AI Act, UK sector regulators, UAE frameworks, ISO 42001. Not a legal textbook. A plain-language answer to four questions: what applies, what it requires, when it kicks in, and what happens if you ignore it.
A board paper you can submit
The Board Paper Template lets you make the business case for AI governance investment in a format the board will read and the audit committee will understand.
Twenty practical governance tools
Every chapter includes at least one tool, template, or checklist — all downloadable. An AI inventory, a risk register, a shadow AI response playbook, a vendor assessment, an agent governance framework, and more.
A sector-specific governance blueprint
Dedicated chapters for financial services, healthcare, professional services, and public sector. Each one addresses the exact AI deployments, regulatory pressures, and liability questions relevant to your industry.
A 90-day implementation roadmap
Week by week. Starting from no dedicated governance team and limited budget. Phase 1 gets you defensible. Phase 2 gets you structured. Phase 3 gets you audit-ready and board-reportable.
One architecture. Every AI system you operate.
The A.C.E. Framework is the book's proprietary governance architecture. It does not change across AI types. What changes is how each layer is operationalised — whether you are governing a credit scoring model, a Microsoft 365 Copilot deployment, or an autonomous claims processing agent.
Strategy, Policy and Inventory
Governance begins before deployment. Connect AI to organisational purpose. Establish the AI inventory, foundation model policy, acceptable use policy, and shadow AI discovery. This is the foundation everything else rests on.
Risk, Oversight and Guardrails
Risk assessment, human oversight design, technical guardrails, copilot governance, RAG governance, vendor management, and the shadow AI response playbook.
Documentation, Monitoring and Audit
You cannot reproduce a specific LLM output. Traditional audit trails do not work for non-deterministic systems. EVIDENCE addresses this with a documentation architecture mapped to ISO 42001 and the EU AI Act.
17 chapters. 4 parts.
One implementation programme.
Built around a deliberate structure: understand your exposure, make the business case, apply the framework, implement it in your sector, and launch your governance programme in 90 days.
The Tuesday Morning Test
A composite scenario. The morning it all surfaces at once.The Case for Governing AI Now
Why governance, why you, why now — and how to secure board buy-in.The A.C.E. Framework
The architecture of AI governance — what to build, in what order.Governing AI in Your Sector
Sector-specific guidance for regulated industries.Building Your Governance Programme
From framework to operating reality.Trusted Intelligence
The standard you set today determines the trust you earn tomorrow.Every chapter ships with tools you can use the same week.
Not decorative templates. Practical instruments sized for mid-market teams — all downloadable from the companion website.
From a 15-question AI Exposure Diagnostic to a week-by-week 90-Day Implementation Roadmap. Every tool is designed to be usable without external guidance.
I have never governed AI at one of the world's largest companies. That is exactly why this book needed to be written.
- ›CEO, Enterprise SaaS Startup
- ›Responsible AI Director, Aligne
- ›Co-founder, Altrum AI Platform
- ›IBM Subject Matter Expert, six years
- ›ISO 42001 Implementor Certified
- ›AIGP, IAPP
- ›CIPP/E, IAPP
- ›AI Ethics and Governance, Oxford Saïd
- ›Operating across UK, EU, UAE
Gurpreet Singh Dhindsa has spent three years sitting across the table from the organisations that the big-firm governance leaders have never spoken to. Mid-market businesses in regulated industries, deploying AI with real commercial pressure, lean teams, and no clear playbook. He has seen what governance failure looks like before it reaches the regulator, and what it costs when it does.
His background is deliberately cross-disciplinary. As CEO of an enterprise SaaS startup, Responsible AI Director at Aligne, and co-founder of the Altrum AI governance platform, he operates across financial services, healthcare, professional services, and public sector clients in London, Dubai, and India. He is not a theorist presenting frameworks designed in a vacuum. Every tool in this book has been tested in practice.
His credentials span the technical, the legal and ethical, and the commercial. That combination is rare in AI governance. It is what makes the advice in this book usable, not just credible.
Be the first to know when it launches.
Join the waiting list for early access, launch-week pricing, exclusive governance resources, and chapter previews before publication.