You know how much your team uses AI. This tells you whether it's actually working.
Most organisations measure AI adoption. Kaivant-O measures AI integration: whether it is genuinely changing how your organisation works, and whether the human capability underneath those AI-assisted results is holding up.
AI output and human capability are not the same thing. The Kaivant Score measures both, and a strong result on one cannot hide a weak result on the other.
An organisation generating strong AI-assisted output on a thinning base of human capability is not performing well. It is building a fragility that today's output metrics cannot see. The score is designed so that reality shows up directly.
Is AI genuinely changing how you work, or just which tools you use?
Leverage Architecture measures whether AI integration has actually reshaped how your organisation coordinates, decides, and delivers work. Tool adoption is not leverage. Leverage is what happens when AI changes the underlying structure: information reaches the right people faster, decisions are made at the right level with better data, and work moves through the organisation with less friction at every step.
Coordination Efficiency
How much has AI reduced the time and effort it takes to get the right information to the right people? High scores mean the organisation is genuinely better connected. Low scores mean more tools, same friction.
Decision Velocity
Are decisions faster because people have better information, or because the analysis is being skipped? Speed built on clarity is the target. Speed built on shortcuts is the failure mode. The instrument distinguishes between the two.
Autonomous Workflow
What proportion of work now moves without a person needing to hand it off and re-explain it at every step? This measures structural change in how work flows, not individual task shortcuts.
Leverage Trajectory
Is the leverage you have built compounding, plateauing, or starting to reverse? This is the forward signal: not where you are today, but what the next twelve months look like from here.
What is happening to the people doing the work?
This is the measurement most organisations are missing. AI can deliver strong output while the human capability underneath that output quietly erodes. People who stop doing the thinking stay competent on paper, until they are suddenly required to work without the tools. Organisational Capital measures whether your people are getting stronger alongside your AI capability, or more dependent on it.
Learning Velocity
Is your organisation actually learning from what it does, or running experiments that leave no lasting change? High Learning Velocity means what people discover gets captured and built on. Low means the organisation keeps re-learning the same things without moving forward.
Human Judgment Utilisation
How much of your people's working time involves genuine thinking, rather than reviewing what AI has already produced? The primary risk of extensive AI deployment is not errors in AI output. It is the gradual loss of the human capacity to catch them.
Capability Development Velocity
Are your people's skills growing as the nature of their work changes? Human capability depreciates when what AI can do expands faster than what people are learning. This dimension measures whether your people are keeping pace, or falling behind without anyone noticing yet.
Human Dignity and Agency
Do your people feel safe raising concerns about AI outputs? Can they push back? Do they have enough ownership over their work to take responsibility for it? An organisation where the answer is no has removed the layer that catches AI errors before they become organisational crises.
Below a defined threshold on Human Dignity and Agency, the Kaivant Score is flagged regardless of how strong the other dimensions are. No leverage result offsets a workforce that has lost the safety, agency, and ownership that allows errors to be caught and performance to be sustained.
What a Kaivant-O result looks like.
The profile below uses fictional scores to show the structure of a Kaivant-O output: two composite axis scores, nine dimension readings each with a current and a forward indicator, and the gap between them, which is the most actionable thing the instrument produces.
Strong demonstrated past adaptation · thinning experimentation pipeline · declining proactive redesign rate. The intervention point is the lead reading, not the lag. This is the AA signal that most directly warrants action.
The leverage this organisation has built is real. Coordination is working well, decisions are happening faster, and a meaningful share of work moves without manual handoffs at every step. These are genuine gains. The problem is on the other side: the human capability that makes those gains durable has not grown at the same pace. The organisation is performing well on AI output while its ability to sustain and build on that output is under pressure.
The leverage trajectory and adaptation architecture readings are telling the same story at different timescales. Adaptation Architecture is the mechanism that generates each new round of improvement: running experiments, learning from them, redesigning how work gets done. When that engine slows, Leverage Trajectory is where the effect shows up first, because the next wave of performance depends on the previous improvement cycle having run.
Both are pointing the same way: the improvement cycle is slowing. Fewer experiments are being run. What is being learned is not being converted into something the organisation can build on. The organisation is producing good current results while quietly drawing down the engine that generated them.
The one dimension moving in the right direction is how much time people are spending on genuine thinking rather than reviewing AI outputs. That is exactly what a recovery needs. But the two dimensions that would build on it, capability development and learning capture, are both declining on their forward indicators. The judgment is there. The conditions to build on it are not yet in place.
The starting point is already there: the judgment being exercised is the raw material. A focused investment in connecting that judgment to deliberate skill-building and systematic learning capture would address both weakening forward indicators at once. Those two dimensions feed directly into the adaptation engine. When that engine is running, the leverage trajectory stabilises. Not as a direct target, but as the natural result of an improvement cycle that is producing each new round of gains.
Is your organisation learning from AI, or just using it?
Adaptation Architecture is the bridge between your leverage output and your human capital. It measures whether your organisation has the capacity to keep improving: running experiments, learning from them, and redesigning how you work so that each cycle builds on the last. Without a functioning adaptation engine, leverage plateaus and eventually reverses.
The most important signal it produces is the gap between your current score and your forward reading. A strong current score with a declining forward reading means the organisation has adapted well in the past and has slowed its active improvement rate. That is the pattern that most directly warrants action: not because today's performance is suffering, but because the mechanism that generated it is weakening.
Each adaptation cycle creates the conditions for the next one. When the engine slows, the compounding stops. Organisations that measure this early enough can intervene before the performance consequences become visible.
A score tells you where you are. The gap tells you where you're going.
Each of the nine dimensions produces two readings: a current score (what is true now) and a forward indicator (where the leading signals point). The gap between them is the most useful thing the instrument produces. A high current score with declining forward indicators means your organisation is performing well today and building the conditions for underperformance in twelve months.
Each score maps to KAIVANT's intervention library: practical, evidence-grounded responses to each pattern the instrument identifies. The score is the starting point. The development path from it is the output.
Reassessment at six and twelve months tracks whether the trajectory has changed. The instrument is designed for longitudinal use: not a one-time audit, but a regular read on whether your organisation is moving in the right direction and whether the actions you have taken are producing the results you need.
Built on serious measurement science.
The Kaivant Score is grounded in established research on expertise development, capability atrophy, and how human-AI integration plays out in practice. The two-axis model, the non-compensatory composite, and the augmentation-substitution framework are anchored in peer-reviewed theory, applied and refined through the instrument design process.
The full theoretical foundation, including the measurement specifications, dimension architecture, and the research basis for each design decision, is published on kaivant.org. The Foundation Paper sets out the conceptual architecture behind the Kaivant Score in full.
The organisations measuring this now will understand something their competitors don't.
Kaivant-O launches on kaivantscore.com in 2026. Register now to be among the first to see where your organisation actually stands: the leverage it has built, and the human capability it is building on.