Governing Intelligence in the AI Era
How organisations direct, verify, and remain accountable for intelligence — human and artificial.
How organisations direct, verify, and remain accountable for intelligence — human and artificial.
Artificial intelligence has entered boardrooms, finance functions, and operating decisions faster than the institutions meant to oversee it. Capability is accelerating; accountability, verification, and judgement are not keeping pace.
Intelligence Governance Architecture is the deliberate design of the structures, roles, disciplines, and controls through which an organisation directs, verifies, and takes responsibility for the intelligence it uses. The Intelligence Governance Architect (IGA) is a practitioner identity, not a job title: a capability waiting to be recognised, developed, and institutionally positioned, whether held by one person or distributed across several.
The core principle: intelligence may be distributed, but accountability may not. A responsible human remains the final arbitrator. This site is being built as a home for this emerging profession and its practitioners.
In practice, the IGA capability helps an organisation know where AI is being used, decide who approves and answers for each decision, record how judgement was exercised at the time, and escalate exceptions before they become losses.
.jpg/:/cr=t:0%25,l:0%25,w:100%25,h:100%25)
How intelligence is organised: roles, reporting lines, decision rights, and controls.
How people and systems actually act: incentives, biases, and the gap between policy and practice.
The norms, rules, and bodies that confer legitimacy and enforce accountability.
How risks and effects change as intelligence is amplified across functions, markets, and societies.
A manufacturing company uses an AI tool to rank suppliers and recommend where to place orders. The tool is fast and usually right. But who is accountable when it is wrong?
Decision rights are written into the Delegation of Authority. The AI recommends; a named purchase manager approves orders up to a set limit, and higher values go to the next level. Someone is assigned to check that supplier data is correct and complete.
Where the purchase team disagrees with the AI, it may override the recommendation, but must record why. Overrides are reviewed periodically, both to correct the tool and to ensure the team is not simply accepting every recommendation without thought.
Each decision leaves a contemporaneous record: the data the AI used, its recommendation, the person who approved or overrode it, and the reasoning at the time, not reconstructed afterwards. Internal audit reviews a sample; the board receives an exception summary.
When the tool is extended from one plant to many, thresholds, monitoring, and escalation are set before rollout, so that a single flaw does not repeat silently across the group.
If a decision causes a loss, the organisation can show who decided, on what information, and why. Intelligence was distributed; accountability stayed with the named people.
Applied together, the four dimensions reveal where governance holds, where it is weak, and where it is absent. The example above shows governance holding. Remove any one element (the named approver, the recorded override, the contemporaneous record, the thresholds set before rollout) and the gap becomes visible: the diagnostic below sets each element of the example against its weak and absent forms.
Holds: A named manager approves AI-recommended orders up to a set limit.
Weak: The Delegation of Authority exists but does not cover AI-recommended orders.
Absent: No one is named; "the system decided".
Holds: Overrides are allowed, reasons are recorded, and overrides are reviewed.
Weak: Overrides are allowed, but reasons are not recorded.
Absent: Every recommendation is accepted without thought.
Holds: A contemporaneous record, an audit sample, and a board exception summary.
Weak: The record is reconstructed after the event.
Absent: There is no record at all.
Holds: Thresholds and monitoring are set before rollout.
Weak: Thresholds are set but not monitored.
Absent: The tool is rolled out to all plants unchanged.

Structure, Scale, and the Architecture of Existence. A philosophical inquiry across scales, from spacetime and quantum fields to the emergence of life, intelligence, and meaning. Notion Press, 2026.

A Structural Inquiry into Technology, Leadership, and Human Responsibility. As AI makes intelligence abundant, the challenge shifts to organising, governing, and using it responsibly. Notion Press, 2026.

In production.
Built on five pillars (Intelligence, Institutions, Judgement, Power, and Discipline) with Emotional Intelligence as their foundation: thirteen chapters, nineteen tools, and four worked scenarios that carry the practitioner from diagnosis and design to execution.
In preparation.
The books form a single line of inquiry: from understanding, to design, to practice.
An Inquiry into the Nature of Reality asks: how does existence organise itself into structure?
Intelligence, Institutions, and Judgement in the AI Era asks: how do human societies organise intelligence?
Governing Intelligence: A Practitioner's Handbook for the Intelligence Governance Architect asks: how does one practitioner govern intelligence honestly, inside one institution?
Understanding does not obligate design. Design does not guarantee practice. Each step down in scale is a step towards greater accountability, not lesser ambition — the cosmic question can be contemplated; the grounded one must be lived, inside an actual institution, by an actual practitioner, on an actual Tuesday.
Management Consultant, Board Director, and Intelligence Governance Architect, based in Navi Mumbai, with five decades of cross-industry experience anchored in finance, governance, and corporate advisory, including senior leadership roles in Indian industry.
Bringing that experience to the defining governance question of the age: how institutions should direct, verify, and remain accountable for intelligence they did not create and cannot fully inspect.
An Inquiry into the Nature of Reality
Intelligence, Institutions, and Judgement in the AI Era
Earlier: Handbook of Factory Accounts
Grasp the structural essence of any manifestation — a career, an institution, a civilisation — and its mastery becomes possible.
Technical AI controls tell an organisation what its models are doing. This work asks who answers for what the organisation does with them: who decided, on what information, and why.
SBIS examines Structure, Behaviour, Institution, and Scale together, so that gaps are found where they actually arise, not only where tools can see them.
Five decades in finance and accounts. Head of a finance, compliance, and accounting academy whose programmes also served governance, risk, compliance, and audit professionals, and for a period responsible for a dedicated GRCA academy as well. From 2012 to 2016, guided the development of e-learning content tailored to the organisation’s needs, with internal and external subject experts. Director on the boards of several companies, attending every meeting and reviewing draft minutes with care, so that the record matched what was decided.
One step outside the system, where independent judgement is most needed and least available. Advice is independent of any technology vendor, with no software, platform, or product to sell.
Advanced AI systems serve as structured thinking partners, to sharpen inquiry, test coherence, and refine articulation, while full authorship of judgement and interpretation stays human. The discipline the framework asks of organisations is applied here first. Client information remains confidential: names and identifying details are removed before any material is used with AI tools, and settings that limit retention and training are used wherever available.
AI and intelligence governance oversight for boards and committees.
An SBIS-based assessment of an organisation's intelligence architecture.
Market entry, feasibility, and corporate strategy.
Sessions for leadership teams, professional bodies, and academic institutions.
A conversation with the board or management to understand where AI is used and what concerns them most.
Examination of selected AI applications, decision rights, delegations, and existing controls.
Assessment of Structure, Behaviour, Institution, and Scale to identify where governance holds, where it is weak, and where it is absent.
A clear report covering gaps found, responsibilities clarified, controls recommended, and a practical plan for implementation.
Engagements may also begin with a workshop for directors or leadership teams.
For advisory engagements, speaking invitations, or enquiries about the books, please write using this form or the chat.
Navi Mumbai, Maharashtra, India
We use cookies to analyze website traffic and optimize your website experience. By accepting our use of cookies, your data will be aggregated with all other user data.