AI Readiness Assessment

Quiz Questions

• 1.
  Your team has adopted Claude in your workflow. You're running 15+ conversations daily—work, personal research, project sketches, client feedback synthesis. At end of week, you can't remember which decision came from where. How do you stay organized?

  • A.
    I just accept the cognitive chaos — that's part of using AI
  • B.
    I try to take notes after each AI conversation, but I often forget
  • C.
    I have a manual system (notes, folder naming, decision log) that I maintain
  • D.
    I use software or AI systematically to trace every decision back to its source
• 2.
  You're juggling email, Slack, Claude chat, Google Docs, and your actual work. Adding an AI co-pilot to the mix feels like one more thing to track. How do you manage this fragmentation?

  • A.
    I don't actively manage it — I just switch between tools as needed
  • B.
    I'm aware it's fragmented, but I haven't found a way to organize it yet
  • C.
    I have some routines to compartmentalize AI work from other tasks
  • D.
    I have clear boundaries and dedicated workflows that prevent tools from overlapping
• 3.
  You've started using AI for multiple work processes simultaneously—drafting, research, brainstorming, summarizing. By Friday, you're mentally exhausted. What is your plan moving forward?

  • A.
    This is just how it is; productivity comes at a mental cost
  • B.
    I notice the burnout but can't figure out how to pace myself differently
  • C.
    I'm getting better at spacing out AI-heavy work across the week
  • D.
    I structure my week to balance AI-intensive tasks with solo/offline work
• 4.
  A vendor offers an AI tool to automate your hiring screening. It's 70% faster. You don't know how the model works, but it's 'industry standard.' What do you do?

  • A.
    Speed wins — I delegate to the tool and stop worrying about it
  • B.
    I'm hesitant, but I feel pressure to keep up with what others are doing
  • C.
    I run a pilot to see if the outputs match my hiring instincts before fully delegating
  • D.
    I audit the model's decisions against my criteria and keep human review on high-stakes calls
• 5.
  Your team wants to use AI to draft customer service responses. It's good 80% of the time, but 20% sounds tone-deaf or slightly inaccurate. How do you handle this risk?

  • A.
    80% is good enough — let it run and adapt as needed
  • B.
    The 20% risk bothers me, but I'm not sure how to solve it
  • C.
    I set up spot-checks on outgoing responses before they go live
  • D.
    I've defined clear boundaries: AI drafts low-risk responses only; humans handle relationship-critical messages
• 6.
  Your manager asks you to use AI to prioritize your team's project backlog. The AI recommends a completely different order than you expected. How do you respond?

  • A.
    I accept the AI's ranking — it probably sees patterns I don't
  • B.
    I'm skeptical but don't know enough to push back confidently
  • C.
    I compare the AI's reasoning to my domain knowledge and adjust if needed
  • D.
    I ask for the AI's criteria, weight them against team capacity and business goals, and make a hybrid decision
• 7.
  You've tried Claude or ChatGPT a few times, and you're mostly copy-pasting basic prompts or having open-ended conversations. A colleague is getting impressive outputs and you want to know how. How do you improve?

  • A.
    I'm interested but don't know where to start
  • B.
    I log that I need to do this on a task list to follow up later and try to google it or ask ChatGPT a week later
  • C.
    I read some prompting tips and experiment with more specific instructions
  • D.
    I actively study prompt structures, test variations, and refine based on output quality
• 8.
  Your company just adopted a new AI tool and it's different from what you've used. You need to use it for a project next week. What's your approach?

  • A.
    I'll wing it when I get to it; they're all basically the same anyway
  • B.
    I know I should learn it, but I feel overwhelmed by the setup
  • C.
    I watch a tutorial and do a quick test before using it for real work
  • D.
    I spend time understanding the tool's strengths, limitations, and specific use cases before diving into my project
• 9.
  You've been using the same AI tool the same way for months. You see a new feature released that might be relevant. What do you do?

  • A.
    New features are probably not worth my time
  • B.
    I notice them but don't feel motivated to change my workflow
  • C.
    I'll try it out if it clearly solves a problem I'm having
  • D.
    I regularly explore new features and evaluate whether they improve my process
• 10.
  AI capabilities are changing monthly. You learned prompting best-practices 3 months ago, but new features make your old approach outdated. How do you stay current?

  • A.
    I stick with what I know — there's always something new anyway
  • B.
    If someone recommends a product I will try it, but things are happening so fast it feels exhausting
  • C.
    I check in periodically to see what's changed and try to keep up
  • D.
    I actively scan announcements and set aside time to experiment with new capabilities
• 11.
  Your company is experimenting with a new AI model that works differently than the previous one. What's your initial reaction?

  • A.
    I'm frustrated by the change — I just figured out the old one
  • B.
    I'm reluctant to relearn, but I know I have to eventually
  • C.
    I'll adapt when I have to
  • D.
    I'm curious about the differences and actively exploring how this model thinks differently
• 12.
  A new category of AI tool emerges (e.g., AI video, AI data analysis) that could help your work, but it requires learning a completely new interface. How do you decide whether to learn it?

  • A.
    I'll skip it — I'm already managing enough tools
  • B.
    It looks useful, but the learning curve is too steep right now
  • C.
    I'll learn it eventually if it becomes necessary for my job
  • D.
    I will try it out, even if there's no immediate pressure or need to do so
• 13.
  Your company has no AI policy. You're considering pasting customer data into Claude to analyze patterns. What's your reasoning?

  • A.
    If it's useful for work, yes
  • B.
    I'm uncomfortable with this
  • C.
    I limit what I share or try to anonymize the data as much as possible
  • D.
    I maintain an inventory of what data is safe to share and use alternative methods when unsure
• 14.
  A colleague casually uses a free AI tool for work. You're not sure if company data is being logged or used to train the model. What do you do?

  • A.
    It's probably fine
  • B.
    Maybe it's an issue, but we don't have a process for reporting this kind of stuff
  • C.
    I ask my colleague questions to learn about what they're doing and make a judgment call
  • D.
    I research the tool's data practices and discuss data safety norms with my team
• 15.
  You accidentally paste sensitive information into an AI tool before realizing it. What happens next?

  • A.
    Nothing I can do now; I move on
  • B.
    I'm worried but don't know who to tell or what to do
  • C.
    I flag it to my manager or IT and ask what the next steps are
  • D.
    I immediately document what was shared, notify the appropriate team, and adjust my practices to prevent recurrence
• 16.
  Your teammate wants to use an AI tool to draft something that affects your shared work. You're not sure it's the right call. How do you handle it?

  • A.
    It's their call as to what products they use
  • B.
    I'm uncertain whether to voice concerns or just go with it
  • C.
    I inquire about the decision and share my perspective
  • D.
    I collaborate to evaluate the tradeoffs, make the decision together, and document the reasoning
• 17.
  A non-technical colleague asks you to explain what you're using AI for. They seem confused or skeptical. How do you respond?

  • A.
    I don't try to explain — they wouldn't understand anyway
  • B.
    I give a quick explanation but get impatient if they have follow-up questions
  • C.
    I try to explain in plain language and answer their questions
  • D.
    I take time to understand their concerns and explain in terms relevant to their work
• 18.
  Your team is divided on whether to adopt a particular AI tool. Some are enthusiastic, others are hesitant. What's your role?

  • A.
    I don't have any expertise here so go with the flow
  • B.
    I ask questions but am no expert and prefer to defer to those who have deeper knowledge of the topic
  • C.
    I listen to both sides and share what I've learned
  • D.
    I facilitate conversation to surface concerns, share my point of view, and build shared understanding before the team decides
• 19.
  Does your organization have documented guidelines about AI use at work?

  • A.
    Yes, clear policies exist
  • B.
    Somewhat — there are guidelines but they're not always clear
  • C.
    Not really — we're still figuring it out
  • D.
    I'm not sure
• 20.
  You want to use an AI tool for a project. How do you navigate your company's AI policies?

  • A.
    I use the tools I want and don't check the policy
  • B.
    I vaguely remember there's a policy but don't reference it
  • C.
    I reference the policy and chat with my team to understand what's approved and follow the guidelines
  • D.
    I understand the policy deeply and advocate for sensible changes when policies are outdated
• 21.
  Your company hasn't set AI policies yet. How do you make decisions about what's safe to do with AI?

  • A.
    I figure it out as I go
  • B.
    I try to be thoughtful, but I'm often unsure if I'm making the right call
  • C.
    I ask my manager or team what they think is acceptable if I am unsure
  • D.
    I establish my own principles based on an understanding of data risk and company culture, and help shape team norms
• 22.
  You notice your company's approach to AI is very different from what you're learning elsewhere (online, in your industry). What do you do?

  • A.
    Our way is the only way that matters for my job
  • B.
    I notice the gap but don't understand what the gap is or how to close it
  • C.
    I share articles or learnings with my manager when relevant
  • D.
    I understand both contexts, help translate best practices to our situation, and advocate for evolution

Quiz Outcomes

• 1.
  Cognitive Load Management Result 1

  The volume AI creates has outpaced the system you have for tracking it, which means decisions are getting made and context is getting lost in ways that are hard to see until something goes wrong. The shift happens when you treat orientation — knowing what's in progress, what got decided, and why — as part of the work, not a cleanup task.Cognitive Load ManagementAI generates volume, context, and decisions at a pace that outstrips most people's ability to track them. The work isn't doing the task; it's knowing what got done, why a decision was made, and what needs to happen next when you have 15 open conversations across three tools with no single source of truth. This category measures whether you've built the habits, systems, and structure to stay oriented when AI amplifies your output without amplifying your capacity to manage it.When this is a gap, it usually shows up in ways that get misattributed. The decisions feel half-remembered, the work feels messier than it used to even though you're technically doing more, and the burnout doesn't respond to working harder because the problem isn't effort — it's the absence of a system that keeps pace with the volume AI creates. People switch tools. The issue is almost never the tools.For teams and organizations, the risk moves from personal to structural here, and that shift is often invisible until something breaks. When cognitive load isn't managed, institutional knowledge lives in someone's chat history, critical context disappears when someone's out, and accountability becomes impossible to trace because no one agreed on what got decided or why. The organizations that build for this — shared logs, decision trails, clean handoff practices — are more resilient, extract compounding value from AI, and can actually trace what happened and why.
• 2.
  Judgment vs Delegation Result 1

  You know the question exists — when to trust the output, when to check it, when to override — but you're not yet applying a consistent framework to answer it. The gap usually shows up as either over-reliance on outputs you haven't verified, or under-use because nothing feels certain enough to act on.Judgment vs DelegationAI will tell you what to do in a way that sounds incredibly convincing, and sometimes will be wrong. The question is whether you should listen. This category tracks the skill that's hardest to teach and most expensive to lose: knowing when to act on an AI output directly, when to check it, and when to override it entirely. This judgment layer is confused with intuition but in reality is practiced, context-dependent, and constantly being tested.A high score here doesn't necessarily mean skepticism of AI. It means using it faster and more accurately because you trust your ability to catch what it gets wrong. A low score usually goes one of two ways: people who delegate too freely and inherit errors they don't notice, or people who are so uncertain they can't benefit from the tool at all. Neither is “safe.”At the team and organizational level, this is where AI risk actually lives. Not in the tool itself — in the chain of humans who touched the output without questioning it. One person’s unchecked AI output becomes a team decision, which becomes a product, which becomes a policy. This category is the signal that organizations with real AI risk exposure should be watching most closely.
• 3.
  Tool Fluency Result 1

  You're using AI, but mostly on its terms — taking what it gives you and working around the gaps rather than diagnosing them. The thing that changes your results fastest isn't learning more tools; it's getting deliberate about why a prompt worked or didn't, and building that feedback loop into how you work.Tool FluencyMost people who think they're using AI well are actually using it casually — copy-pasting prompts, accepting first outputs, and never really testing whether there's a better approach. Tool fluency isn't about how many tools you've tried; it's about whether you understand them well enough to use them with actual intent. That means knowing how the same prompt performs differently across models, recognizing when output quality is degrading and why, and knowing when the problem is the tool and when the problem is the prompt.Individually, this is what separates people getting similar outputs from AI and people extracting dramatically different value from the same tools. Someone with high fluency knows which model to use and why, catches quality degradation early, and makes accurate calls about when to put the tool down. Someone with low fluency works harder with the same tools, gets worse outputs, and usually can't identify which part of the process is the problem.For teams and organizations, uneven fluency is one of the most common and most invisible problems in AI adoption. The people running laps and the people barely engaged rarely discuss it — so the gap compounds. Best practices don't spread, the organization believes it's “using AI” while capturing a fraction of what's available, and by the time it becomes visible it's already a culture problem as much as a skills problem. Collective fluency is the thing that actually moves an organization forward.
• 4.
  Adaptive Learning Result 1

  Your current approach was probably built for a more stable environment, and AI is changing fast enough that some of what you learned six months ago is already working against you. The shift isn't about learning more — it's about building the habit of updating, which looks like treating your current practices as provisional rather than settled.Adaptive LearningAI capabilities are evolving faster than any onboarding program or annual training cycle can track, which means the relevant question isn't what you know right now — it's how you behave when what you know becomes outdated. New models, new tool categories, and new norms emerge constantly, and best practices from six months ago are sometimes actively counterproductive today. This category measures whether you've built your approach to work around the assumption that the ground will keep moving.Individually, this shows up as how you behave when your usual approach stops working — whether you experiment and update or get frustrated and wait. People who score low here aren't unintelligent; they've often invested heavily in learning a particular set of practices, and that investment creates resistance to changing them even when the evidence suggests they should. The skills you've built will still erode faster than you expect if you treat them as fixed, because the environment they were built for keeps shifting.For organizations, this is the difference between teams that build compounding advantage and teams that are always chasing the last announcement. Organizations that cultivate adaptive learning — where people share what they're discovering, experiment deliberately, and normalize updating their practices — develop institutional responsiveness that gets faster over time. Teams without it repeat the same adoption cycle: roll out a tool, train people on a fixed curriculum, watch the training become obsolete, repeat — each round more expensive and more demoralizing than the last.
• 5.
  Team/Collaborative Readiness Result 1

  Your AI practice is mostly individual, which works until the work touches other people — and then the gaps in shared standards, inconsistent outputs, and unspoken assumptions start creating friction that's hard to trace back to its source. The shift starts with making your own process visible, not just your outputs.Team & Collaborative ReadinessMost AI adoption frameworks are written as if individuals are adopting AI, when in reality people are adopting AI inside organizations where their colleagues have different tools, different comfort levels, different risk tolerances, and often completely incompatible assumptions about what responsible use even looks like. This category measures whether you can operate effectively in that environment — sharing what you know, creating space for people who are behind, raising concerns about practices that create shared risk.Individually, a high score here means you're doing the work with awareness of the people around you — you notice when colleagues are struggling and create space for that, you explain your reasoning and your process, not just hand over the output. And when you see team-level practices that create risk — inconsistent standards, unchecked outputs going into shared work — you say something, even when it's not your job and no one asked. That behavior is rarer than it sounds and more valuable than most organizations recognize.For organizations, low collaborative readiness produces a specific kind of fragmentation that's easy to miss: everyone is using AI, but no one is using it the same way, no one knows what standards apply, and the gap between the most and least capable people keeps widening without anyone naming it. The result is misalignment that compounds — inconsistent outputs, duplicated effort, a creeping inability to agree on what good work looks like, and individual capability that stays siloed rather than building into something shared. Organizations with high collaborative readiness build shared standards faster, surface problems earlier, and get organizational value from the capability that already exists on their teams.
• 6.
  Organizational Context Awareness Result 1

  You know your organization has policies and positions on AI, but you're operating mostly on general best practices rather than a specific read of your actual context. The gap shows up in edge cases — when the right move isn't clear from policy and you're guessing rather than reading the room.Organizational Context AwarenessGeneric AI best practices are written for a generic organization that doesn't exist. Real organizations have specific cultures, specific risk profiles, specific leadership positions on AI that may be partially formed or internally inconsistent, and specific gaps between what policy says and what people actually do. This category measures whether you understand the organization you're actually in well enough to act with good judgment inside it — with enough judgment to act well in the space between what policy says and what the situation actually requires.High performance here isn't compliance — it's something harder and more useful. It's the ability to read an ambiguous situation inside a specific organization and make a judgment call that's defensible, appropriate to the actual risk level, and doesn't require waiting for someone else to write a clearer rule. People who score well here know what their organization has and hasn't sanctioned, understand where the real gray areas live versus where people just think there are gray areas, and raise concerns in ways that move the organization's norms forward.At the organizational level, this is what separates workforces that can operate with genuine autonomy from workforces that are dependent on clear instructions that rarely arrive. When contextual awareness is low across an organization, what looks like AI adoption is actually supervised compliance — people follow rules when they exist and guess badly when they don't. When it's high, people make defensible judgment calls in the gap between policy and reality, catch problems before they escalate, and contribute to norms as they form. That's the version of AI-ready that's actually hard to build and nearly impossible to fake.
• 7.
  Privacy & Security Result 2

  Your instincts here are probably better than your habits — you know there are lines, but you're not consistently asking the question before you cross them. The practice that changes this is simple but deliberate: making data judgment a step in the workflow rather than something you think about after.Privacy & SecurityEvery time you paste something into an AI system, you're making a decision about what leaves your organization's control — and most people aren't making that decision at all, they're just getting the task done. The risk isn't malicious behavior; it's the absence of habitual, practiced judgment about what should and shouldn't cross that line. This category measures whether you've developed that judgment, whether you'd recognize a problem before it became one, and whether your defaults actually protect you or just feel like they do.Individually, this reflects how you operate in the gap between what your organization has formally sanctioned and what you're doing day-to-day — a gap that is almost always wider than people think. Someone with high readiness here has developed their own working principles, knows what kinds of data should and shouldn't cross the line, asks that question before they paste, and treats data judgment as part of doing the work — not a compliance concern parked with someone else.For teams and organizations, this is where liability actually lives — not in rogue actors but in ordinary, well-intentioned decisions made without adequate judgment. The pattern is almost always the same: an employee uses a tool that seems fine, shares data that seems fine, in a context that seems fine, until it isn't. Organizations that manage this well build a workforce where that judgment is practiced and automatic, and that's the version of security that actually scales.
• 8.
  Team/Collaborative Readiness Result 2

  You're aware of the team dimension and operating with some consideration for it, but you're probably not yet raising concerns about shared practices or closing the gap for colleagues who are behind. That next step is the difference between individual readiness and organizational contribution.Team & Collaborative ReadinessMost AI adoption frameworks are written as if individuals are adopting AI, when in reality people are adopting AI inside organizations where their colleagues have different tools, different comfort levels, different risk tolerances, and often completely incompatible assumptions about what responsible use even looks like. This category measures whether you can operate effectively in that environment — sharing what you know, creating space for people who are behind, raising concerns about practices that create shared risk.Individually, a high score here means you're doing the work with awareness of the people around you — you notice when colleagues are struggling and create space for that, you explain your reasoning and your process, not just hand over the output. And when you see team-level practices that create risk — inconsistent standards, unchecked outputs going into shared work — you say something, even when it's not your job and no one asked. That behavior is rarer than it sounds and more valuable than most organizations recognize.For organizations, low collaborative readiness produces a specific kind of fragmentation that's easy to miss: everyone is using AI, but no one is using it the same way, no one knows what standards apply, and the gap between the most and least capable people keeps widening without anyone naming it. The result is misalignment that compounds — inconsistent outputs, duplicated effort, a creeping inability to agree on what good work looks like, and individual capability that stays siloed rather than building into something shared. Organizations with high collaborative readiness build shared standards faster, surface problems earlier, and get organizational value from the capability that already exists on their teams.
• 9.
  Team/Collaborative Readiness Result 3

  You're doing the work that makes AI adoption actually functional at the team level — sharing process, flagging risk, creating space for the people who are still figuring it out. That behavior is rarer than it sounds, and it's the thing that determines whether an organization gets compounding value from AI or just inconsistent individual results.Team & Collaborative ReadinessMost AI adoption frameworks are written as if individuals are adopting AI, when in reality people are adopting AI inside organizations where their colleagues have different tools, different comfort levels, different risk tolerances, and often completely incompatible assumptions about what responsible use even looks like. This category measures whether you can operate effectively in that environment — sharing what you know, creating space for people who are behind, raising concerns about practices that create shared risk.Individually, a high score here means you're doing the work with awareness of the people around you — you notice when colleagues are struggling and create space for that, you explain your reasoning and your process, not just hand over the output. And when you see team-level practices that create risk — inconsistent standards, unchecked outputs going into shared work — you say something, even when it's not your job and no one asked. That behavior is rarer than it sounds and more valuable than most organizations recognize.For organizations, low collaborative readiness produces a specific kind of fragmentation that's easy to miss: everyone is using AI, but no one is using it the same way, no one knows what standards apply, and the gap between the most and least capable people keeps widening without anyone naming it. The result is misalignment that compounds — inconsistent outputs, duplicated effort, a creeping inability to agree on what good work looks like, and individual capability that stays siloed rather than building into something shared. Organizations with high collaborative readiness build shared standards faster, surface problems earlier, and get organizational value from the capability that already exists on their teams.
• 10.
  Cognitive Load Management Result 2

  You have enough structure to stay oriented most of the time, but it's probably fragile — dependent on your own memory or habits that don't hold up when volume spikes or collaborators are involved. Tightening the handoff layer is where the next level of reliability comes from.Cognitive Load ManagementAI generates volume, context, and decisions at a pace that outstrips most people's ability to track them. The work isn't doing the task; it's knowing what got done, why a decision was made, and what needs to happen next when you have 15 open conversations across three tools with no single source of truth. This category measures whether you've built the habits, systems, and structure to stay oriented when AI amplifies your output without amplifying your capacity to manage it.When this is a gap, it usually shows up in ways that get misattributed. The decisions feel half-remembered, the work feels messier than it used to even though you're technically doing more, and the burnout doesn't respond to working harder because the problem isn't effort — it's the absence of a system that keeps pace with the volume AI creates. People switch tools. The issue is almost never the tools.For teams and organizations, the risk moves from personal to structural here, and that shift is often invisible until something breaks. When cognitive load isn't managed, institutional knowledge lives in someone's chat history, critical context disappears when someone's out, and accountability becomes impossible to trace because no one agreed on what got decided or why. The organizations that build for this — shared logs, decision trails, clean handoff practices — are more resilient, extract compounding value from AI, and can actually trace what happened and why.
• 11.
  Cognitive Load Management Result 3

  You've built the habits and systems that let AI amplify output without creating chaos, and you're one of the people keeping institutional knowledge from living exclusively in chat histories. That's rarer than it sounds and worth protecting as your team's AI use scales.Cognitive Load ManagementAI generates volume, context, and decisions at a pace that outstrips most people's ability to track them. The work isn't doing the task; it's knowing what got done, why a decision was made, and what needs to happen next when you have 15 open conversations across three tools with no single source of truth. This category measures whether you've built the habits, systems, and structure to stay oriented when AI amplifies your output without amplifying your capacity to manage it.When this is a gap, it usually shows up in ways that get misattributed. The decisions feel half-remembered, the work feels messier than it used to even though you're technically doing more, and the burnout doesn't respond to working harder because the problem isn't effort — it's the absence of a system that keeps pace with the volume AI creates. People switch tools. The issue is almost never the tools.For teams and organizations, the risk moves from personal to structural here, and that shift is often invisible until something breaks. When cognitive load isn't managed, institutional knowledge lives in someone's chat history, critical context disappears when someone's out, and accountability becomes impossible to trace because no one agreed on what got decided or why. The organizations that build for this — shared logs, decision trails, clean handoff practices — are more resilient, extract compounding value from AI, and can actually trace what happened and why.
• 12.
  Judgment vs Delegation Result 2

  You're making judgment calls actively rather than defaulting in either direction, which puts you ahead of most. Where this gets sharper is in the consistency — applying the same scrutiny to AI outputs that feel right as the ones that feel off.Judgment vs DelegationAI will tell you what to do in a way that sounds incredibly convincing, and sometimes will be wrong. The question is whether you should listen. This category tracks the skill that's hardest to teach and most expensive to lose: knowing when to act on an AI output directly, when to check it, and when to override it entirely. This judgment layer is confused with intuition but in reality is practiced, context-dependent, and constantly being tested.A high score here doesn't necessarily mean skepticism of AI. It means using it faster and more accurately because you trust your ability to catch what it gets wrong. A low score usually goes one of two ways: people who delegate too freely and inherit errors they don't notice, or people who are so uncertain they can't benefit from the tool at all. Neither is “safe.”At the team and organizational level, this is where AI risk actually lives. Not in the tool itself — in the chain of humans who touched the output without questioning it. One person’s unchecked AI output becomes a team decision, which becomes a product, which becomes a policy. This category is the signal that organizations with real AI risk exposure should be watching most closely.
• 13.
  Judgment vs Delegation Result 3

  You've internalized the judgment layer to the point where it's part of how you work, not a separate step. That makes you significantly more reliable with AI than people who score similarly on fluency but skip the verification — and it makes the outputs you touch more trustworthy downstream.Judgment vs DelegationAI will tell you what to do in a way that sounds incredibly convincing, and sometimes will be wrong. The question is whether you should listen. This category tracks the skill that's hardest to teach and most expensive to lose: knowing when to act on an AI output directly, when to check it, and when to override it entirely. This judgment layer is confused with intuition but in reality is practiced, context-dependent, and constantly being tested.A high score here doesn't necessarily mean skepticism of AI. It means using it faster and more accurately because you trust your ability to catch what it gets wrong. A low score usually goes one of two ways: people who delegate too freely and inherit errors they don't notice, or people who are so uncertain they can't benefit from the tool at all. Neither is “safe.”At the team and organizational level, this is where AI risk actually lives. Not in the tool itself — in the chain of humans who touched the output without questioning it. One person’s unchecked AI output becomes a team decision, which becomes a product, which becomes a policy. This category is the signal that organizations with real AI risk exposure should be watching most closely.
• 14.
  Tool Fluency Result 2

  You understand your tools well enough to get consistent value from them, and you can tell the difference between a tool problem and a prompt problem most of the time. The next edge comes from testing more systematically — treating outputs as hypotheses rather than drafts.Tool FluencyMost people who think they're using AI well are actually using it casually — copy-pasting prompts, accepting first outputs, and never really testing whether there's a better approach. Tool fluency isn't about how many tools you've tried; it's about whether you understand them well enough to use them with actual intent. That means knowing how the same prompt performs differently across models, recognizing when output quality is degrading and why, and knowing when the problem is the tool and when the problem is the prompt.Individually, this is what separates people getting similar outputs from AI and people extracting dramatically different value from the same tools. Someone with high fluency knows which model to use and why, catches quality degradation early, and makes accurate calls about when to put the tool down. Someone with low fluency works harder with the same tools, gets worse outputs, and usually can't identify which part of the process is the problem.For teams and organizations, uneven fluency is one of the most common and most invisible problems in AI adoption. The people running laps and the people barely engaged rarely discuss it — so the gap compounds. Best practices don't spread, the organization believes it's “using AI” while capturing a fraction of what's available, and by the time it becomes visible it's already a culture problem as much as a skills problem. Collective fluency is the thing that actually moves an organization forward.
• 15.
  Tool Fluency Result 3

  You're operating at a level most people in your organization aren't, and you probably know it because you can see the gap between what you're getting and what colleagues assume is possible. The leverage at this level is in building that fluency into shared practices, not just your own workflow.Tool FluencyMost people who think they're using AI well are actually using it casually — copy-pasting prompts, accepting first outputs, and never really testing whether there's a better approach. Tool fluency isn't about how many tools you've tried; it's about whether you understand them well enough to use them with actual intent. That means knowing how the same prompt performs differently across models, recognizing when output quality is degrading and why, and knowing when the problem is the tool and when the problem is the prompt.Individually, this is what separates people getting similar outputs from AI and people extracting dramatically different value from the same tools. Someone with high fluency knows which model to use and why, catches quality degradation early, and makes accurate calls about when to put the tool down. Someone with low fluency works harder with the same tools, gets worse outputs, and usually can't identify which part of the process is the problem.For teams and organizations, uneven fluency is one of the most common and most invisible problems in AI adoption. The people running laps and the people barely engaged rarely discuss it — so the gap compounds. Best practices don't spread, the organization believes it's “using AI” while capturing a fraction of what's available, and by the time it becomes visible it's already a culture problem as much as a skills problem. Collective fluency is the thing that actually moves an organization forward.
• 16.
  Adaptive Learning Result 2

  You update when you have to, and you're paying enough attention to notice when things shift. What separates this tier from the next is whether updating is reactive or deliberate — whether you're waiting for friction to prompt a change or actively scanning for when your approach has gotten stale.Adaptive LearningAI capabilities are evolving faster than any onboarding program or annual training cycle can track, which means the relevant question isn't what you know right now — it's how you behave when what you know becomes outdated. New models, new tool categories, and new norms emerge constantly, and best practices from six months ago are sometimes actively counterproductive today. This category measures whether you've built your approach to work around the assumption that the ground will keep moving.Individually, this shows up as how you behave when your usual approach stops working — whether you experiment and update or get frustrated and wait. People who score low here aren't unintelligent; they've often invested heavily in learning a particular set of practices, and that investment creates resistance to changing them even when the evidence suggests they should. The skills you've built will still erode faster than you expect if you treat them as fixed, because the environment they were built for keeps shifting.For organizations, this is the difference between teams that build compounding advantage and teams that are always chasing the last announcement. Organizations that cultivate adaptive learning — where people share what they're discovering, experiment deliberately, and normalize updating their practices — develop institutional responsiveness that gets faster over time. Teams without it repeat the same adoption cycle: roll out a tool, train people on a fixed curriculum, watch the training become obsolete, repeat — each round more expensive and more demoralizing than the last.
• 17.
  Adaptive Learning Result 3

  You've built your approach around the assumption that the ground keeps moving, which means you compound rather than reset when the landscape changes. That institutional responsiveness is hard to develop and easy to lose — it's worth making it a named practice on any team you're part of.Adaptive LearningAI capabilities are evolving faster than any onboarding program or annual training cycle can track, which means the relevant question isn't what you know right now — it's how you behave when what you know becomes outdated. New models, new tool categories, and new norms emerge constantly, and best practices from six months ago are sometimes actively counterproductive today. This category measures whether you've built your approach to work around the assumption that the ground will keep moving.Individually, this shows up as how you behave when your usual approach stops working — whether you experiment and update or get frustrated and wait. People who score low here aren't unintelligent; they've often invested heavily in learning a particular set of practices, and that investment creates resistance to changing them even when the evidence suggests they should. The skills you've built will still erode faster than you expect if you treat them as fixed, because the environment they were built for keeps shifting.For organizations, this is the difference between teams that build compounding advantage and teams that are always chasing the last announcement. Organizations that cultivate adaptive learning — where people share what they're discovering, experiment deliberately, and normalize updating their practices — develop institutional responsiveness that gets faster over time. Teams without it repeat the same adoption cycle: roll out a tool, train people on a fixed curriculum, watch the training become obsolete, repeat — each round more expensive and more demoralizing than the last.
• 18.
  Privacy & Security Result 3

  You're operating with a working sense of what should and shouldn't cross the line, and you're applying it in practice rather than just knowing it in theory. Where this gets sharper is in the edge cases — the data that seems fine in isolation but creates risk in context, or the tool that's approved but not for this use.Privacy & SecurityEvery time you paste something into an AI system, you're making a decision about what leaves your organization's control — and most people aren't making that decision at all, they're just getting the task done. The risk isn't malicious behavior; it's the absence of habitual, practiced judgment about what should and shouldn't cross that line. This category measures whether you've developed that judgment, whether you'd recognize a problem before it became one, and whether your defaults actually protect you or just feel like they do.Individually, this reflects how you operate in the gap between what your organization has formally sanctioned and what you're doing day-to-day — a gap that is almost always wider than people think. Someone with high readiness here has developed their own working principles, knows what kinds of data should and shouldn't cross the line, asks that question before they paste, and treats data judgment as part of doing the work — not a compliance concern parked with someone else.For teams and organizations, this is where liability actually lives — not in rogue actors but in ordinary, well-intentioned decisions made without adequate judgment. The pattern is almost always the same: an employee uses a tool that seems fine, shares data that seems fine, in a context that seems fine, until it isn't. Organizations that manage this well build a workforce where that judgment is practiced and automatic, and that's the version of security that actually scales.
• 19.
  Privacy & Security Result 4

  You've developed the kind of practiced data judgment that most organizations are trying to write policies to enforce. That means you're protecting your organization in ways it probably can't see, and you're doing it as part of the work rather than as compliance overhead.Privacy & SecurityEvery time you paste something into an AI system, you're making a decision about what leaves your organization's control — and most people aren't making that decision at all, they're just getting the task done. The risk isn't malicious behavior; it's the absence of habitual, practiced judgment about what should and shouldn't cross that line. This category measures whether you've developed that judgment, whether you'd recognize a problem before it became one, and whether your defaults actually protect you or just feel like they do.Individually, this reflects how you operate in the gap between what your organization has formally sanctioned and what you're doing day-to-day — a gap that is almost always wider than people think. Someone with high readiness here has developed their own working principles, knows what kinds of data should and shouldn't cross the line, asks that question before they paste, and treats data judgment as part of doing the work — not a compliance concern parked with someone else.For teams and organizations, this is where liability actually lives — not in rogue actors but in ordinary, well-intentioned decisions made without adequate judgment. The pattern is almost always the same: an employee uses a tool that seems fine, shares data that seems fine, in a context that seems fine, until it isn't. Organizations that manage this well build a workforce where that judgment is practiced and automatic, and that's the version of security that actually scales.
• 20.
  Organizational Context Awareness Result 2

  You have a working sense of your organization's actual stance on AI — not just the formal policy, but the real risk tolerance, the gray areas, and where the norms are still forming. Where this gets sharper is in using that read proactively, not just to avoid problems but to help shape how the norms develop.Organizational Context AwarenessGeneric AI best practices are written for a generic organization that doesn't exist. Real organizations have specific cultures, specific risk profiles, specific leadership positions on AI that may be partially formed or internally inconsistent, and specific gaps between what policy says and what people actually do. This category measures whether you understand the organization you're actually in well enough to act with good judgment inside it — with enough judgment to act well in the space between what policy says and what the situation actually requires.High performance here isn't compliance — it's something harder and more useful. It's the ability to read an ambiguous situation inside a specific organization and make a judgment call that's defensible, appropriate to the actual risk level, and doesn't require waiting for someone else to write a clearer rule. People who score well here know what their organization has and hasn't sanctioned, understand where the real gray areas live versus where people just think there are gray areas, and raise concerns in ways that move the organization's norms forward.At the organizational level, this is what separates workforces that can operate with genuine autonomy from workforces that are dependent on clear instructions that rarely arrive. When contextual awareness is low across an organization, what looks like AI adoption is actually supervised compliance — people follow rules when they exist and guess badly when they don't. When it's high, people make defensible judgment calls in the gap between policy and reality, catch problems before they escalate, and contribute to norms as they form. That's the version of AI-ready that's actually hard to build and nearly impossible to fake.
• 21.
  Organizational Context Awareness Result 3

  You understand your organization's AI context at the level where you can operate with genuine autonomy — making defensible judgment calls in the gaps between policy and reality without waiting for someone to write a clearer rule. That's the version of AI-ready that's actually hard to build, and it makes you a multiplier for the people around you.Organizational Context AwarenessGeneric AI best practices are written for a generic organization that doesn't exist. Real organizations have specific cultures, specific risk profiles, specific leadership positions on AI that may be partially formed or internally inconsistent, and specific gaps between what policy says and what people actually do. This category measures whether you understand the organization you're actually in well enough to act with good judgment inside it — with enough judgment to act well in the space between what policy says and what the situation actually requires.High performance here isn't compliance — it's something harder and more useful. It's the ability to read an ambiguous situation inside a specific organization and make a judgment call that's defensible, appropriate to the actual risk level, and doesn't require waiting for someone else to write a clearer rule. People who score well here know what their organization has and hasn't sanctioned, understand where the real gray areas live versus where people just think there are gray areas, and raise concerns in ways that move the organization's norms forward.At the organizational level, this is what separates workforces that can operate with genuine autonomy from workforces that are dependent on clear instructions that rarely arrive. When contextual awareness is low across an organization, what looks like AI adoption is actually supervised compliance — people follow rules when they exist and guess badly when they don't. When it's high, people make defensible judgment calls in the gap between policy and reality, catch problems before they escalate, and contribute to norms as they form. That's the version of AI-ready that's actually hard to build and nearly impossible to fake.
• 22.
  Overall Result 3

  Understand what your readiness looks like across each of the seven dimensions of AI Readiness below.Over the next several weeks, I'll be doing a deep dive into how people and teams can level up their AI Readiness across each of these dimensions. Follow along by subscribing to The Business Side of AI.