AI Competency | Essay 4
What Comes First: AI Competency or AI Strategy?
Essay 3 laid out four foundational investments: unified data, scalable compute, organization-wide AI fluency, and functional governance, and argued that these create the conditions for AI competency required to compete in the agentic AI era. This essay makes a stronger claim: those investments don’t just precede strategy, they produce it.
A company that unifies its data discovers opportunities invisible when that data was siloed, and a team that develops AI fluency identifies opportunities no external consultant could, because they know where the operational friction actually lives. A governance framework built around real systems clarifies risk boundaries no pre-deployment policy document could anticipate.
This is not an argument against strategic thinking, it is an argument against strategic waiting. Legacy firms do not need an AI strategy, they need a business strategy with AI competency embedded as a foundational investment. Not a separate initiative expected to produce ROI in a vacuum within a few months, but a permanent investment similar to how they invest in financial competency, operational competency, and legal competency, so the organization is well positioned to use AI in shaping decisions and actions continuously.
The Case Against AI Strategy
When a business labels something an “AI strategy,” it has already separated AI from the business. It becomes a parallel workstream running alongside operations rather than through them.
This is how most legacy organizations operate today. Business strategy is owned by the CEO and the board. AI strategy is owned by a CAIO or CIO, with its own roadmap, budget line, and success metrics. The two are loosely connected by use cases that promise to drive value, but in practice they run on different timelines, report to different stakeholders, and compete for the same resources.
The result is, AI strategy produces pilots while the business strategy produces revenue and the two never fully integrate. The pilots either die when funding cycles shift or survive as isolated tools that never reach the root of how the company actually competes.
If your AI strategy sounds like “transform customer experience” or “optimize operational processes,” the framing is wrong. AI can certainly improve experience, deliver productivity gains, and drive operational efficiency, however, calling that a strategy mistakes an AI project for a durable competency.
An AI strategy asks: where can we apply AI to improve the business? This produces a ranked list of use cases evaluated by feasibility and ROI, which the organization funds and executes as a program or initiative. These initiatives produce activity, and activity produces outputs without any durable way to convert outputs to outcomes. When the projects ship, the strategy is declared complete, but the underlying business, how it makes decisions, how it competes, how it compounds advantage, remains unchanged.
A business strategy asks a fundamentally different question: what is our core competency, what decisions define it, and how does AI compound our mastery of those decisions over time? This does not produce a temporary initiative. It produces a permanent orientation, a way of operating where every function continuously improves its decision-making through AI, the way every function continuously improves its financial discipline through accounting.
The distinction between an AI strategy and AI competency embedded in a business strategy is not semantic. It determines whether AI investment is directed at delivering outputs or at compounding the core competencies that drive sustainable business outcomes.
Three Advantages of Competency Framing
It eliminates the false separation between AI and the business. When AI competency is embedded in the business strategy, there is no parallel roadmap to maintain, no separate budget to defend, and no translation layer between what the AI team builds and what the business needs. The investment in AI is the investment in the business because they are the same thing, and every dollar spent on AI infrastructure is evaluated against the same question the board already asks: does this strengthen our competitive position?
It forces alignment with core business competency. An AI strategy can fund anything: an HR chatbot, a marketing content generator, an internal meeting summarizer. Any use case that produces ROI looks like progress. A business strategy with an embedded AI competency applies a harder filter: does this investment compound our mastery of the decisions that define our market position? If a logistics company is spending its AI budget on internal productivity tools instead of a proprietary routing engine, the competency framing exposes that misalignment immediately. The AI strategy framing never asks the question.
It survives leadership transitions. AI strategies are almost always tied to the executive who championed them, and when that person leaves the strategy stalls. In the current market, CAIO tenure is measured in months rather than years, which makes this a structural vulnerability rather than an edge case. AI competency does not depend on a single sponsor because it is woven into how the organization operates. Finance does not collapse when the CFO departs and operations do not stall when the COO transitions. AI competency should carry the same institutional durability.
With this framing, the starting point for AI competency investment is not a use case audit. It is a decision audit.
The Decision Audit
For decades, business strategy was built by mining external data: market research, competitive benchmarking, and consultant frameworks built on cross-industry pattern matching. The strategic intelligence that mattered most lived outside the organization, and the firms that could access and interpret it fastest held the advantage. The AI economy inverts this.
The most valuable strategic intelligence is now internal, living in the proprietary data your organization has been generating for years, in the transaction patterns, the supply chain decisions, the customer behavior, and the operational friction points that no external benchmarking study can see.
LLMs and agentic coding now allow even the least technically mature legacy firm to convert proprietary knowledge into growth opportunities at a pace that was not possible before. The firms that will define the next era are the ones that learn to mine their own operational data for setting the strategic direction in their domain rather than waiting for a consulting firm to tell them where the market is heading.
A decision audit is how that mining begins. Every business runs on a small number of recurring, high-stakes decisions made at enormous volume. Visa asks whether a transaction is fraudulent. Airbnb asks whether a booking is a risk. John Deere asks whether a plant is a weed or a crop. These decisions are made millions of times a day, and each one produces data that sharpens the next. Over time, the accumulated mastery of those decisions becomes uncopyable, not because of the model powering them but because of the decades of decisions the model was trained on and the model’s ability to incorporate every future decision into its next one.
Visa’s fraud detection is not superior because Visa has better algorithms. It is superior because Visa has been making fraud decisions at scale for thirty years, and every transaction has made the next one marginally more accurate. The fraud detection use case does not create the competency. The competency investment creates compounding value from fraud detection. The decision audit precedes competency, competency precedes the use case, and that is the order of operations.
A decision audit asks which decisions the organization wants to keep getting better at, forever. This question cannot be answered by a CAIO working in isolation or a strategy team ranking use cases or running discovery exercises across the organization. It requires the CEO and board, because it is a question about what the business is and what it could evolve to, not what AI can do. If you are working on a list of AI projects without first conducting a decision audit, the likelihood of delivering durable business value is very low.
Where to Begin
The answer to the question that paralyzes most leadership teams, where do we start?, is not a use case. It is a decision.
Most enterprises already operate with a moments that matter framework for customer experience, identifying the critical touchpoints where the interaction builds or erodes loyalty. The strategic equivalent for the business itself is a decisions that matter framework: a deliberate inventory of the recurring, high-stakes decisions that define your competitive position, paired with a commitment to compounding the organization’s mastery of them.
Identify the two or three recurring decisions that most define your competitive position. Then ask honestly: what data would make those decisions better, is that data unified and accessible, is there governance for the systems that will act on it, and are there people across the organization who can pivot from hindsight-based workflows to ones designed around what happens next?
The answers will tell you exactly where to begin, not because a strategy document prescribed it, but because the decisions that matter to your business revealed it.
Thank you for reading my essay.
Essays 1 through 4 conclude Part 1 of my three-part series on AI Competency. I will be on a pause during the summer months, and upon my return Part 2 will dive into what AI competency is in practice and how businesses can invest in developing it. Part 3 will cover measuring progress, outcomes, and implications for the workforce and the future of work.
If you found these essays insightful, please share your thoughts in comments.
Previous essays on AI Competency
AI Competency Essay 1 - Core Competency of a Business - An Introduction
AI Competency Essay 2 - Business Core Competency as a Guide to Your AI Implementation Strategy
AI Competency Essay 3 - Creating the Conditions for AI Competency
About the Author: Meghna Sinha is Chief AI Officer and Co-founder of Kai Roses, Inc. Kai Roses helps organizations build AI competency that compounds. Contact Meghna at www.kairoses.com to discuss building AI competency for your organization.