AI, HUMAN INTELLIGENCE AND THE GOVERNANCE GAP

A Discussion Document

Compiled from an extended strategic dialogue  |  May 2026

This document summarises an extended strategic dialogue covering artificial intelligence, geopolitical risk, the limitations of machine cognition, and the design of a transformation program for business leaders. The discussion was initiated by a senior commercial practitioner with direct operational experience across four geopolitical theatres spanning three decades. It is presented here in the third person as a coherent record of the themes, concerns, and proposed solutions that emerged.

WHO: THE PRINCIPAL AND HIS CONTEXT

The discussion was led by a senior commercial practitioner — referred to throughout this document as the Principal — with an operational career spanning four decades and direct experience in some of the most contested commercial environments of the late twentieth and early twenty-first centuries.

The Principal's distinctive credential is not academic or theoretical. Between 1987 and 2018, the Principal was subjected to four separate incidents of unlawful detention across four geopolitical theatres: Nigeria (1987), Russia (1991), Mexico (1999), and the Iraq-Kuwait border region (2018). Two of these incidents were subsequently acknowledged by Chinese state actors as state-directed operations. In each case, extraction was achieved autonomously, without institutional or governmental support. Subsequently assassination attempts were made in the United Kingdom, Africa and the Middle East in regard to these and four more UK operations.

These experiences are not offered as biography. They are the evidence base for a body of knowledge — concerning how state and criminal actors assess, target, and manage commercial individuals — that is directly relevant to the challenges organisations now face in the AI era. The Principal now coaches others and is developing a program to transmit this knowledge to senior business leaders.

The question isn't why it kept happening to the same individual. The question is why it doesn't happen to more people — and the answer is that most companies never send anyone effective enough into those environments to be worth targeting in the first place.

WHAT: THE CORE THEMES OF THE DISCUSSION

1. The AI Transition as a Civilisational Event

The discussion opened with an assessment of artificial intelligence not as a technology trend but as a civilisational shift — one that will fundamentally alter the distribution of power, the nature of competition, and the cognitive habits of human beings. Senator Bernie Sanders' warnings about AI (particularly Mythos) were noted as a useful frame: his concerns about robotic warfare, mass surveillance, and the concentration of extraordinary control in the hands of small elites are not fringe concerns. They describe trajectories already visible in multiple state actors.

The consensus of the discussion was that AI will deliver systemic shock to every existing political and economic order simultaneously — capitalism in the West, authoritarian models in Russia and China, and everything in between. None of these systems has a graceful adaptation pathway. The question is not whether disruption will occur but whether the humans responsible for navigating it will have the cognitive and relational capacities to do so wisely.

2. The Geopolitical Landscape: A New Map

The discussion examined the emerging AI geopolitical order in detail. The traditional bipolar framing — United States versus China — was assessed as insufficient. A more accurate picture involves five distinct actors whose interests and capabilities create a genuinely multipolar landscape:

— The United States remains the dominant AI capability power but is increasingly transactional in its treatment of allies, and its AI leadership is partly dependent on Gulf state infrastructure.

— China operates a sophisticated commercial intelligence apparatus and has demonstrated willingness to take direct action against commercial actors perceived as threatening its interests — as the Principal's own experience confirms.

— The Gulf States — particularly the UAE and Saudi Arabia — are emerging as a structural third pole. With sovereign wealth funds managing close to six trillion dollars and the UAE alone having committed $148 billion to AI since early 2024, these states are building the infrastructure layer upon which both Western and Eastern AI systems will increasingly depend. Their strategic interest is indispensability, not alignment.

— Russia represents an asymmetric AI threat — not hyperscale ambition but weaponised disruption through disinformation, deepfake operations, and cyber capabilities deployed against democratic infrastructure.

— Non-state actors, including ISIS-affiliated groups, are already exploring AI for recruitment, communications, and weapons modification, representing a diffuse threat layer that no conventional diplomatic framework adequately addresses.

The implications for British foreign policy were examined in detail. The discussion concluded that the United Kingdom's most distinctive strategic asset in this landscape is its potential role as honest broker of AI governance — a position established at the Bletchley AI Safety Summit but requiring active investment to maintain. Britain must treat AI infrastructure as foreign policy, appoint technically literate AI ambassadors, and resist the binary framing of US versus China in favour of a genuinely multipolar strategy.

3. The Cognitive Atrophy Problem

A central theme of the discussion was the risk that AI adoption, if unmanaged, will erode precisely the human capacities most needed to govern it wisely. This was described not as a concern about IQ but about the atrophy of discernment — the ability to read a room, sense when something is wrong before data confirms it, understand motivation beyond what algorithms can model, and adapt to genuinely unprecedented situations.

The Principal's operational experience across four decades and four theatres represents an extreme version of the capability at stake. The ability to assess threat signals in novel environments, extract meaning from incomplete information, and make consequential decisions without institutional support — these are not natural gifts. They are developed through experience, reflection, and exposure to genuinely high-stakes situations.

The concern raised is that AI-dependent organisations, by optimising for data-driven decision-making, are actively weakening the human capacity for environmental reading at precisely the moment when the environment is becoming maximally complex and adversarial.

AI systems excel at processing known data within modelled parameters. They cannot read a room. They cannot detect the moment a commercial counterpart shifts from partner to adversary. These are human cognitive capabilities — and they are precisely the capabilities that state actors, criminal networks, and sophisticated commercial adversaries exploit when they engage with AI-dependent organisations.

4. The Academic Speed Problem

The discussion addressed the structural mismatch between the pace of AI development and the pace of academic knowledge dissemination. Peer review cycles of six to eighteen months mean that published research describes a reality that may already be obsolete. Several solutions were identified: the formalisation of preprint servers such as arXiv as primary academic literature, the development of rapid-review journals with genuine institutional prestige, living documents updated on version-control systems, and AI-assisted peer review that compresses timelines without removing human judgement.

The deeper problem identified was one of incentive structures: until universities and funding bodies formally reward rapid, open, and iterative publication, researchers will remain constrained by systems designed for a slower world.

5. The Governance Gap

The discussion identified three specific failures in how AI-dependent organisations are currently governed:

Direction without context: Direction without context.

AI systems are directed by those who define their parameters. Where the people providing that direction have no experiential understanding of how adversaries behave in contested commercial environments, the AI will be optimised for a world that does not exist.

Governance without adversarial literacy: Governance without adversarial literacy.

Governance frameworks for AI focus predominantly on ethics, bias, and regulatory compliance. Almost none address how a sophisticated state or criminal actor will probe, test, and exploit an AI-dependent commercial operation.

Management without human signal reading: Management without human signal reading.

Senior management in AI-first organisations increasingly rely on AI outputs for commercial decision-making. The human capacity to detect early-stage threat signals atrophies through disuse, precisely when it is most needed.

WHERE AND WHEN: THE CONTEXT OF URGENCY

The discussion was not abstract. The geopolitical and commercial pressures described are live and accelerating. The Gulf states are committing hundreds of billions to AI infrastructure now. China's commercial intelligence operations are active now. Western AI capabilities are dependent on chip supply chains, energy infrastructure, and data centre capacity that are already subjects of geopolitical contest.

The adaptation shock to existing economic and political systems is not a future event. It is occurring in real time, and the organisations and leaders who will navigate it most effectively are those who begin building the relevant capacities now rather than waiting for the disruption to force adaptation under duress.

Britain's position is particularly acute. Post-Brexit, the United Kingdom sits outside the EU regulatory framework and must navigate its relationship with the United States, China, and the Gulf states simultaneously, with diplomatic tools designed for a slower and simpler world. The window in which its Bletchley-established governance credibility represents genuine first-mover advantage is finite.

WHY: THE CASE FOR URGENT ACTION

The discussion identified several convergent reasons why the issues raised require urgent rather than deferred attention:

— The compounding nature of AI development means that governance deficits do not remain stable — they widen. An organisation or state that lacks adversarial literacy today will be proportionally more vulnerable next year.

— The geopolitical landscape is hardening into configurations that will be difficult to renegotiate once established. The Gulf states' infrastructure dependencies, China's Belt and Road AI extensions, and the US regulatory posture on chip exports are all creating facts on the ground.

— The cognitive atrophy of leadership is cumulative. Leaders who have not developed discernment and environmental reading over time cannot acquire it rapidly when a crisis demands it. The time to build that capacity is before the crisis.

— The social and economic shock of AI adoption is accelerating. Organisations that have not prepared their people — not just their systems — for transformation rather than transaction will face breakdown rather than adaptation.

— The window for norm-setting in AI governance is open but will not remain so indefinitely. Nations and organisations that help define the rules of the game now will have structural advantage over those that arrive late.

The AI revolution will happen regardless. The question is whether it happens to businesses and societies, or whether it happens through them — with human wisdom guiding it. That depends on leaders capable of thinking beyond optimisation.

HOW: THE PROPOSED SOLUTION

The Transformation Program

The discussion culminated in the outline of a 24-month leadership transformation program, designed to address the governance gap by building the human capabilities that AI cannot provide. The program is structured around four pillars:

Pillar One: Restored Cognition — The Practice of Discernment

This pillar addresses the atrophy of environmental reading and adversarial thinking in data-dependent leaders. It is delivered through small cohort work with experienced practitioners — deliberately not through case studies, but through engagement with real, current, ambiguous situations that participants are facing in their actual organisations. The Principal's operational experience across four theatres provides the foundational teaching framework.

Pillar Two: Resilience as Strategic Architecture

This pillar reframes resilience from a risk management function to an umbrella organisational strategy. It involves structured audit and redesign of participants' actual organisations — mapping fragility, building in adaptive slack, creating decision structures that function under genuine uncertainty, and understanding the economics of antifragility versus pure efficiency.

Pillar Three: The Trust Foundation — Women as Architects of Adaptive Capacity

The discussion identified relational intelligence — the capacity to build trust, read emotional undercurrents, hold ambiguity, and connect disparate parts of a system — as both systematically undervalued in current leadership culture and critically important for navigating AI's consequences. This pillar explicitly positions women as architects of how organisations learn and sense their environments, not merely as participants in what they produce. The program is designed with explicit gender balance and women in facilitation leadership.

Pillar Four: Geopolitical Literacy and Systemic Shock Preparation

This pillar equips leaders to understand the actual geopolitical landscape in which their organisations operate — including the AI power dynamics between the United States, China, the Gulf states, Russia, and non-state actors. It uses scenario planning with genuine geopolitical complexity, stress-testing participants' business models against multiple difficult futures, and building the capacity to operate when normal institutional assumptions break.

Program Structure

The program is designed as a 24-month cohort experience for 12-15 senior leaders, deliberately mixed across industry, geography, gender, and functional background. It operates through four quarterly immersions of three to four days each, supplemented by monthly peer consultation calls and individual coaching. Physical presence is required — the cognitive and relational learning the program addresses cannot be delivered virtually.

Facilitators combine the Principal's geopolitical and operational expertise with specialists in organisational resilience, women's leadership, and live geopolitical analysis. The program is positioned not as leadership development — a phrase associated with incremental improvement — but as strategic transformation for the AI age, addressing existential business risk.

CONCLUSION: THE LARGER ARGUMENT

The discussion that this document summarises was not a single-topic conversation. It moved from AI geopolitics to diplomatic strategy to personal operational experience to academic reform to leadership education because these are not separate subjects. They are expressions of a single underlying condition: a world changing faster than the human and institutional capacities built to navigate it.

The Principal's four decades of operational experience across contested environments — and the cognitive capacities that experience developed — represent an unusual vantage point from which to assess that condition. The argument that emerges is coherent and urgent: AI amplifies human capability but does not substitute for human discernment. The organisations and nations that will navigate the transition most effectively are those that invest simultaneously in AI capability and in the human capacities that give that capability wisdom and direction.

The governance gap — the failure to recognise that AI deployment without adversarial literacy, relational intelligence, and genuine resilience thinking is dangerous rather than merely incomplete — is the central problem the program outlined in this document exists to address.

The market for this is enormous. Every serious organisation is trying to figure out how to navigate AI without losing its humanity. None of them have a real answer yet. The experience base and program design outlined in this document represents a credible, evidence-grounded response to that need.

— End of Document —