THE SCENARIO
A small island nation in the Caribbean contracted with a global technology corporation to provide an AI-driven public service delivery platform covering digital identity, social benefits distribution, tax filing, business registration, land records, and public records management. The platform was implemented with remarkable speed — the government was under political pressure to demonstrate digital transformation results before the next election cycle, and the provider promised a fully operational digital government within eighteen months. The contract was a standard government services agreement — the same terms the provider offered to dozens of other small nations. No sovereignty assessment was conducted during the procurement process. No testing for local applicability was performed. No data residency commitments were negotiated beyond the provider’s standard language. No exit provisions existed beyond the basic termination-for-cause clause. The platform was implemented on schedule, and the government celebrated its digital transformation success.
Twenty months after deployment, the provider announced a strategic restructuring of its government services division. Pricing tiers changed: the small nation was reclassified from “strategic partner” to “standard customer,” triggering a 40% price increase. Data residency commitments were modified: the provider consolidated regional data centres, moving the nation’s citizen data to a jurisdiction with weaker privacy protections. The feature roadmap was redirected toward larger clients — the custom integrations that the nation had requested were deprioritised indefinitely. The small nation, which had dismantled most of its legacy paper-based systems during the digital transformation, discovered that it could not exit the platform without disrupting essential services to millions of citizens. It had built a digital state on leased land — and the leaseholder had changed the terms without the lessee’s consent. The strategy they had not built was now costing them far more than the strategy they had declined to invest in.
Are you building your technology future on rented land — or do you own the foundation?
Building a strategy for technology dependency requires a fundamental shift in how leaders think about their relationship with technology providers. The dominant approach — adopt first, govern later, or worse, adopt and never govern — has produced the current crisis of unmanaged dependency that institutions around the world are now confronting. The TEE Method™ — Test, Evaluate, Evolve — provides a structured alternative: a framework for building technology strategy that is sovereignty-conscious from the ground up, practical in its implementation, and progressive in its ambition.
This article focuses on the strategic dimension: how leaders move from awareness of dependency to active governance of it, from reactive procurement to deliberate portfolio management, from technology adoption as a series of discrete purchasing decisions to technology governance as an integrated strategic function. The goal is not to eliminate dependency — for most institutions, complete elimination is neither realistic nor necessary. The goal is to ensure that every dependency is chosen, understood, governed, and progressively reduced over time.
Part One: Adoption Without Interrogation — The Central Failure Mode of the AI Era
The Test domain of the TEE Method™ takes aim at the central failure mode of the current technology era: adoption without interrogation. Institutions around the world — governments, corporations, universities, hospitals, non-profits — are deploying AI systems across every domain of activity without asking the most fundamental governance questions. They are adopting platforms without testing them against local conditions. They are signing contracts without reading the sovereignty implications. They are building dependencies without planning for exit. They are, in the TEE Method™’s phrase, adopting without testing — and the testing they skip is not a technical formality but a governance safeguard.
The TEE Method™ insists that every AI system must be tested — not just for technical performance, but for alignment with human flourishing, institutional resilience, and the preservation of sovereignty. Testing, in this framework, is not a pre-deployment checkbox that can be completed once and forgotten. It is a governance discipline that continues throughout the system’s lifecycle, adapting as the system evolves, as the institution’s understanding deepens, and as the external environment changes.
The five domains of TEE Testing form the foundation of any sovereignty-conscious technology strategy:
| Test Domain | Core Question | Governance Implication | Method |
|---|---|---|---|
| 1. Technical Performance | Does the system do what it claims to do, reliably and consistently? | Independent validation against vendor-provided benchmarks, in your context, on your data | Controlled pilot with local data, independent red team evaluation |
| 2. Fairness and Bias | Does the system produce systematically different outcomes for different demographic groups? | Demographic impact assessment across all relevant population segments, not just aggregate metrics | Disparate impact analysis, subgroup performance testing, demographic audit |
| 3. Sovereignty Alignment | Does the system serve your sovereignty interests or undermine them? | Stack position analysis, dependency mapping, data flow documentation, contract term review | TEE Diagnostic Grid™ assessment across all seven stack layers |
| 4. Cultural Compatibility | Does the system respect and accommodate local values, practices, and institutional culture? | Cultural impact assessment before deployment, ongoing monitoring for cultural displacement effects | Cultural audit, stakeholder consultation, values alignment assessment |
| 5. Exit Viability | Can the system be removed, replaced, or exited without causing institutional crisis? | Exit protocol design and testing before deployment, switching cost analysis, alternative readiness assessment | Exit drill, switch cost calculation, alternative provider evaluation |
The Evaluate domain builds on this foundation by asking a second-order question: given that we have tested this system and understand its technical performance, fairness characteristics, sovereignty implications, cultural compatibility, and exit viability — what does this mean for our institution? Evaluation is the bridge between testing and decision-making. It transforms test data into strategic insight, and strategic insight into governance action.
Evaluation operates through the seven dimensions of the Global AI Stack, examining how a system’s performance across the five test domains interacts with the institution’s position at each layer of the stack. A system that performs well on technical benchmarks but creates unacceptable data dependency at the data layer may be acceptable for some institutions and unacceptable for others — evaluation determines which.
Part Two: The Three Categories of Dependency — And What Each Requires
Not all dependencies are equal, and treating them as if they were leads to governance mistakes: either over-investing in managing dependencies that could be tolerated, or under-investing in managing dependencies that represent existential risk. The TEE Method™ distinguishes between three categories of dependency, each requiring a different governance approach, a different resource allocation, and a different strategic posture.
Strategic Dependency. This is dependency that the institution has deliberately chosen, actively governs, and maintains the capacity to exit. The institution understands why it depends on a particular provider, has negotiated terms that protect its sovereignty interests, and has verified its ability to operate without the provider if necessary. Strategic dependency includes explicit exit provisions, data sovereignty protections, knowledge transfer commitments, regular reassessment schedules, and independent verification of the provider’s compliance. Strategic dependency is not a weakness — it is a managed relationship in which the institution retains the power to remain or leave as its interests dictate.
Accretive Dependency. This is dependency that accumulates over time without deliberate choice or conscious awareness. A team adopts a collaboration tool. Another team adopts a compatible communication platform. The IT department standardises on a cloud provider to reduce costs. A vendor bundles an AI feature into an existing product. Each individual decision seems reasonable, even beneficial. Collectively, they produce a stack of dependencies that no one decided on as a whole, that no one has mapped, and that no one can govern. Accretive dependency is the most common form of technology dependency and the most dangerous because it operates below the level of institutional awareness. No one approved the architecture. It just grew — and the growth was invisible because it happened through the aggregation of individually rational decisions that produced a collectively irrational outcome.
Enforced Dependency. This is dependency that the institution cannot exit without severe disruption — because the provider has engineered switching costs into the architecture, because contractual lock-in prevents transition, because institutional capacity has atrophied to the point that the institution cannot function without the provider, or because no viable alternatives exist. Enforced dependency is the governance failure that the TEE Method™ is designed to prevent. Once an institution reaches this state, its options are severely constrained: it can accept the provider’s terms, pay the exorbitant cost of exit, or seek regulatory intervention — none of which is a satisfactory position for a sovereign entity.
The strategic objective of any technology dependency strategy is to convert accretive dependencies into strategic dependencies where feasible, to prevent enforced dependencies from developing, and to maintain a clear-eyed understanding of which dependencies fall into each category at any given time. This requires ongoing monitoring, because dependencies can migrate between categories without the institution’s awareness — a strategic dependency can become an enforced dependency if the provider changes its terms and the institution’s alternatives have not been maintained.
Every technology adoption is a sovereignty transaction. The question is not whether you will have dependencies — it is whether those dependencies will be chosen or accumulated, governed or ignored, reversible or permanent. The most expensive dependency is the one you did not know you had.
TEE Method™
Part Three: The Red Flag Checklist for Strategy Gaps
Use this checklist to assess whether your institution’s technology dependency strategy is adequate to the risks you face. The questions test strategic awareness, not technical knowledge — they can be answered by leadership without engineering expertise. If three or more items apply, your institution is operating with a strategy gap that represents a material risk to institutional sovereignty.
- Your institution has adopted AI systems without conducting independent testing for local applicability — you relied on vendor-provided benchmarks and international certifications.
- You cannot produce a complete inventory of all AI systems in use across your organisation, including systems that individual teams have adopted without central approval.
- Your procurement process does not include a mandatory sovereignty assessment as a gate before any AI contract is signed or renewed.
- You have not calculated the switching costs — financial, operational, reputational, and strategic — associated with any major technology provider in your portfolio.
- Data from your institution flows to third-party AI systems without an explicit data sovereignty agreement that specifies ownership, usage rights, retention periods, and deletion protocols.
- You have no documented and tested exit protocol for any AI system currently in production — no plan for what happens if the provider ceases operations, changes terms, or is acquired by a hostile entity.
- Your leadership has not received a briefing on technology sovereignty risks in the past twelve months, and no regular reporting cycle exists for this topic.
- No individual or governing body in your organisation has been assigned explicit responsibility for monitoring technology dependency and developing sovereignty strategy.
- Your institution treats technology procurement as a series of independent purchasing decisions rather than as a portfolio of strategically managed dependencies.
- You cannot name, from memory and without research, the single AI system whose failure or removal would cause the greatest disruption to your institution’s core mission.
Threshold: If three or more of these conditions describe your institution, you are operating with a strategy gap that requires immediate corrective action. The cost of intervention increases with time — the longer dependency goes ungoverned, the more expensive and disruptive any corrective action becomes, and the more likely the institution is to transition from strategic dependency to enforced dependency without recognising the transition until it is complete.
Part Four: The Strategy-Building Framework
The TEE Method™ provides a four-phase framework for building a technology dependency strategy that is sovereignty-conscious, practical, and progressive. Each phase builds on the previous, and the framework is designed to be iterative — the strategy is revisited and revised as the technology landscape changes and as the institution’s understanding of its dependencies deepens.
Phase 1: Map and Measure (Weeks 1-4). Conduct a comprehensive AI inventory across the entire institution. Identify every system in use, every data flow, every provider relationship, every contractual dependency. Quantify dependencies across the seven layers of the Global AI Stack. This phase answers the question: what do we actually depend on? Most organisations discover that their perceived dependency map — the systems they think they use — bears little resemblance to their actual dependency map — the systems that are actually in use, including those adopted without formal approval, those embedded in larger platforms, and those that have been operating for years below the level of leadership awareness.
Phase 2: Assess and Prioritise (Weeks 5-8). For each identified dependency, assess sovereignty risk using the TEE Diagnostic Grid™. Score across technical, operational, governance, and strategic dimensions. Classify each dependency as strategic, accretive, or enforced. Prioritise dependencies based on three factors: strategic criticality, risk level, and feasibility of intervention. This phase answers the question: which dependencies require immediate attention, and which can be addressed over a longer timeframe?
Phase 3: Design and Negotiate (Weeks 9-12). For high-priority dependencies, develop a governance plan. For existing contracts, this may involve renegotiating terms to include sovereignty protections — data sovereignty clauses, exit provisions, knowledge transfer requirements, independent audit rights. For new procurements, this means building sovereignty requirements into the request for proposal, evaluating providers against sovereignty criteria alongside technical and cost criteria, and selecting providers based on their willingness to meet sovereignty standards. For internally built systems, this means ensuring that they do not create new dependencies through embedded third-party components, proprietary data formats, or single-vendor infrastructure choices.
Phase 4: Implement and Evolve (Ongoing). Execute the governance plan. Monitor compliance with negotiated terms. Conduct regular reassessments — at least annually, and more frequently for high-risk dependencies. Evolve the strategy as the technology landscape changes, as new providers enter the market, as existing providers change their behaviour, and as the institution’s own sovereignty priorities develop. The TEE Method™ does not produce a static strategy document that can be filed and forgotten. It produces a governance practice that adapts continuously — a living strategy that evolves with the institution and its environment.
Action Plan: Building Your Technology Dependency Strategy
The following action plan provides a practical pathway from strategy awareness to strategy implementation. It is designed for leaders who want to move from understanding the problem to executing the solution, with clear deliverables at each stage.
| Timeframe | Action | Deliverable | Accountability |
|---|---|---|---|
| Week 1 | Establish a technology sovereignty working group with representatives from technology, legal, procurement, operations, risk management, and governance functions. Assign an executive sponsor at C-suite or minister level. | Chartered working group with named members, clear terms of reference, and executive mandate | Executive sponsor confirms scope and resources |
| Week 2 | Commission a comprehensive AI inventory across all departments, divisions, and operating units. Include formally adopted systems, informally adopted tools, and AI components embedded in larger platforms. | Complete AI asset register with provider mapping, contract identification, and dependency classification | Working group reviews and validates inventory completeness |
| Month 2 | Complete TEE assessments for the five highest-risk AI dependencies. Score each across all five test domains and all seven stack layers. Document findings, risks, and recommended priority actions. | Five completed TEE Diagnostic Grid™ assessments with dependency scores and action recommendations | Independent reviewer validates assessment quality |
| Month 3 | Present full dependency strategy to institutional leadership with recommended priority interventions, resource requirements, timeline, and accountability framework. Secure formal approval. | Approved technology dependency governance plan with budget allocation, milestones, and monitoring cycle | Leadership adopts plan, allocates resources, schedules first quarterly review |
Part Five: Why Strategy Must Precede Crisis
The reason most institutions lack a technology dependency strategy is not that they do not recognise the importance of one. It is that the consequences of not having a strategy are not immediately visible — they accumulate gradually, through the accretion of individually reasonable decisions, until the institution reaches a point where its strategic options have been foreclosed without its awareness. The technology dependency strategy is not a response to crisis. It is a crisis prevention mechanism.
The institutions that invest in building a strategy before they need it are the ones that will have options when the technology landscape shifts — when providers change their business models, when regulatory requirements evolve, when geopolitical events disrupt supply chains, when the cost of dependency exceeds the benefit. The institutions that wait until the crisis arrives will discover that the time and resources required to build a strategy in crisis are far greater than the time and resources required to build a strategy in advance — and that some options, once foreclosed, cannot be reopened.
The Closing Question
The age of unexamined intelligence does not require you to reject technology. It requires you to examine it before you adopt it, to govern it while you use it, and to maintain the capacity to exist without it. The technology companies that are your current partners will continue to evolve their business models, pricing structures, and strategic priorities in directions that may or may not align with your sovereignty interests. Your dependency strategy must be able to evolve with them — or away from them.
Here is the question that every leader must answer, not in a strategy document that gathers dust but in the governance structures they create and the decisions they make every quarter:
If every technology provider you currently depend on changed their terms tomorrow — pricing doubled, data residency commitments ended, support for your region or sector was discontinued — how long would it take your institution to adapt? What are you doing today, this week, this quarter to shorten that timeline and expand your options?
This article draws on the TEE Method™ framework from SOVEREIGN: Who Owns the Future? by Tonisha Tagoe — a comprehensive framework for understanding, assessing, and governing technology dependency in the age of unexamined intelligence.
The institutions that will succeed in the age of unexamined intelligence are not those that avoid dependency — they are those that understand their dependencies, govern them deliberately, and maintain the strategic capacity to act independently when their interests require it. Strategy is not a document. It is a practice — and the practice of sovereignty-conscious technology governance must be embedded in the institution’s culture, procurement processes, contract management, and leadership accountability structures. Every year that passes without a dependency strategy is a year in which dependency deepens, options narrow, and the cost of intervention rises. The time to build the strategy is now — not when the crisis arrives, but before it does.
The TEE Method™ does not promise a future without dependency. It promises a future in which dependency is chosen, understood, governed, and progressively reduced — a future in which the institution retains the power to decide which dependencies to accept, which to reject, and which to reduce over time. That power is the essence of sovereignty in the age of AI, and it is available to any institution that is willing to do the work of building it.