Am I Good Enough at AI for a Promotion?
Are your AI skills for promotion strong enough? Data shows most professionals overestimate by 17.6 points. Find your real score.
If you're wondering whether your AI skills for promotion are where they need to be, you're asking the right question at the right time. Across 2,578 completed AISA assessments, the average composite score sits at 46.2 out of 100 — firmly in the Developing tier. Most professionals believe they're Proficient. Most aren't. The gap between perception and reality is the single biggest career risk nobody's talking about in performance reviews.
This post breaks down what managers actually look for, what "good enough" means in concrete terms, and how to close the gap before your next review cycle.
The New Promotion Filter: Why AI Fluency Now Matters
AI fluency has become a practical differentiator in promotion decisions because managers need people who can do more with the same headcount. It's not about being an AI researcher — it's about whether you can use AI tools to ship faster, think more clearly, and make better decisions.
A 2024 McKinsey Global Survey found that 72% of organisations had adopted AI in at least one business function, up from 55% the prior year. When adoption is that widespread, the question shifts from "does our team use AI?" to "who on our team uses it well?" That's the filter.
What Managers Actually Evaluate
Promotion committees rarely have a line item labelled "AI skills." Instead, AI fluency shows up indirectly. Managers notice who can:
- Draft a first version of a strategy doc, spec, or analysis in half the time
- Catch errors in AI-generated output before they reach stakeholders
- Choose the right tool for the task instead of defaulting to ChatGPT for everything
- Explain their process — not just the output, but why they trust it
These aren't abstract competencies. They're observable behaviours that separate someone ready for the next level from someone who isn't.
The Shift From "Nice to Have" to "Expected"
Two years ago, using AI at work was a bonus. Now it's table stakes. The World Economic Forum's Future of Jobs Report 2025 projected that 39% of workers' core skills would change by 2030, with AI and big data literacy topping the list of fastest-growing skills. When your VP reads that report — and they have — they start looking at their team differently.
The implication for your career: if you're competing for a promotion against someone with demonstrably stronger AI skills, you need to know where you stand. Not where you think you stand.
What "Good Enough" Looks Like: The Proficient Threshold
The Proficient tier in AISA's scoring framework spans a composite score of 60-79, and it represents the practical threshold where someone consistently applies AI effectively across work contexts. Only 23.5% of the 2,578 people assessed reach this tier or above.
Understanding the AI Fluency Levels
AISA's composite tiers map to real behavioural differences:
| Tier | Score Range | Share of Assessees | What It Means in Practice |
|---|---|---|---|
| Emerging | 0-27 | 23.5% | Minimal engagement with AI tools |
| Developing | 28-59 | 45.7% | Uses AI but inconsistently; gaps in critical evaluation |
| Proficient | 60-79 | 23.5% | Reliable, effective AI use across multiple dimensions |
| Advanced | 80-91 | 6.1% | Designs workflows, mentors others, handles edge cases |
| Expert | 92-100 | 1.1% | Pushes boundaries; deep technical and strategic fluency |
The median composite score across all assessees is 46 — squarely in Developing. That means more than half of professionals who care enough to take an AI assessment still fall short of Proficient.
Why 60 Is the Number That Matters
Proficient isn't perfection. It means you can prompt effectively, evaluate output critically, understand enough about how models work to make good tool choices, integrate AI into real workflows, and handle data responsibly. It's the point where a manager can trust you to use AI without supervision.
For context, even among engineers — the highest-scoring role group — the average composite is 54.1. Product managers average 55. Founders average 55.6. None of these groups clear the Proficient threshold on average. If you're at 60+, you're ahead of most of your peers, regardless of role.
What Proficient Looks Like Day-to-Day
A Proficient professional doesn't just use AI — they use it with confidence calibration. They know when to trust output and when to verify. They can explain to a colleague why a particular prompt structure works better than another. They understand that a model's training data has a cutoff and factor that into their decisions. These are the behaviours that show up in promotion conversations, even when nobody uses the word "fluency."
The Five Dimensions Hiring Managers Care About
AISA measures AI fluency across five dimensions, each weighted to reflect its importance in real work. The relative importance shifts by role, but all five matter everywhere. Here's how they break down, with average scores by role from 2,578 assessments.
Dimension Scores by Role
| Dimension (Weight) | Engineering (n=525) | Product (n=149) | Data (n=57) | Design (n=61) | Founders (n=232) |
|---|---|---|---|---|---|
| Prompting & Communication (23%) | 51 | 53.4 | 46.9 | 47.9 | 52.4 |
| Critical Thinking (22%) | 47.9 | 49.5 | 46.7 | 41.3 | 50.3 |
| Technical Understanding (20%) | 50.3 | 46.9 | 42.6 | 39 | 49.5 |
| Workflow & Application (25%) | 54.8 | 57.2 | 49.7 | 50.1 | 57.3 |
| Safety & Responsibility (10%) | 47.4 | 47.2 | 45.4 | 36.1 | 50.7 |
What Each Dimension Means for Your Promotion Case
Workflow & Application carries the highest weight at 25%, and it's where most role groups score highest. This dimension covers task decomposition — breaking complex work into AI-suitable subtasks — and integrating AI into existing processes. Managers care about this because it directly translates to output and velocity. If you can show you've redesigned a workflow to incorporate AI and it saved measurable time, that's promotion material.
Prompting & Communication (23%) is the most visible dimension. It's what people see when they watch you work with AI. Strong prompting isn't about memorising templates — it's about structuring requests clearly, providing appropriate context, and iterating based on output quality. Product managers tend to score highest here (53.4), likely because structured communication is already core to their role.
Critical Thinking (22%) is the dimension most people underestimate. It covers your ability to evaluate AI output for accuracy, detect hallucinations, and make judgment calls about when AI-generated content is good enough to use. Founders score highest here (50.3), but even they're well below Proficient. If you want to read more about this dimension specifically, see Critical Thinking With AI: How Good Are You?
Technical Understanding (20%) doesn't mean you need to train models. It means understanding concepts like context windows, token limits, and model selection well enough to make informed choices. Engineers lead here (50.3), but designers (39) and data professionals (42.6) often have significant gaps.
Safety & Responsibility (10%) has the lowest weight but can be a dealbreaker. One data privacy incident from careless AI use can end a promotion conversation permanently. Designers score lowest here at 36.1 — a gap worth closing. The free Responsible AI Use guide covers the essentials.

Curious about your AI Fluency?
AISA helps you measure, prove and improve your AI skills — free report in a 20–40 minute chat.
The Overconfidence Trap: Your AI Skills Aren't Where You Think
Across 1,668 assessees who predicted their score before taking AISA, the average predicted score was 62.2. The average actual score was 44.6. That's an overconfidence gap of 17.6 points — enough to place someone in an entirely different tier than they expected.
This isn't a rounding error. It's a systematic pattern, and it varies dramatically by role.
Overconfidence by Role
| Role | Avg Predicted | Avg Actual | Gap |
|---|---|---|---|
| Students (n=155) | 66.3 | 35.1 | 31.2 |
| Data (n=38) | 67.8 | 48.6 | 19.2 |
| Design (n=46) | 62.3 | 46.2 | 16.1 |
| Engineering (n=331) | 68.7 | 53.6 | 15.1 |
| Product (n=83) | 63.5 | 53.8 | 9.7 |
| Founders (n=125) | 61.5 | 54.8 | 6.7 |
Engineers predicted 68.7 — solidly Proficient — but scored 53.6 on average. Designers predicted 62.3 and landed at 46.2. The pattern is consistent: the more confident you feel, the more likely you are to be surprised by your actual score. For a deeper dive into this data, see AI Overconfidence by Role: Who's Worst?
Why This Matters for Promotions
Overconfidence creates a specific career risk. You walk into a review believing your AI skills are a strength. Your manager, who's seen your actual output, may disagree. Without an objective benchmark, this mismatch festers. You don't get the feedback you need because nobody has a shared vocabulary for "you're a 45, not a 65."
The professionals who close this gap fastest are the ones who get an honest baseline first. That's not a pitch — it's an observation from the data. People motivated by professional development average 43.2 on AISA, while those motivated by leadership roles average 51.4. The difference isn't talent; it's that leadership-motivated professionals tend to have already confronted their gaps.
The Dunning-Kruger Dynamic in AI
The overconfidence pattern maps closely to the Dunning-Kruger effect. People with surface-level AI exposure — they've used ChatGPT, maybe tried Copilot — rate themselves highly because they don't know what they don't know. They haven't encountered scenarios that test hallucination detection, multi-step workflow design, or responsible data handling. The assessment surfaces these blind spots.
Founders show the smallest gap (6.7 points), possibly because running a company forces you to confront the limits of your tools daily. Students show the largest gap (31.2 points), which makes sense — they've had less opportunity to test AI skills against real-world consequences.
5 Steps to Close the Gap Before Your Next Review
Knowing the gap exists is step one. Closing it requires deliberate practice across specific dimensions. Here's a concrete plan.
1. Get Your Baseline Score
You can't improve what you can't measure. Take the AISA assessment — it's a conversation, not a multiple-choice quiz, and it produces a detailed breakdown across all five dimensions. The published rubric shows exactly what each score level means, so there are no surprises about how you're evaluated. Your report will show you precisely where you're strong and where you're not.
2. Identify Your Weakest Dimension and Focus There First
Don't try to improve everything at once. Look at your dimension scores and find the lowest one. For most people, that's Technical Understanding (population average: 39.2) or Safety & Responsibility (population average: 41.6). Spend two weeks focused on that single dimension.
If Technical Understanding is your gap, start with the How AI Works guide. If it's prompting, work through the Prompting Skills guide. These are free and designed to build practical fluency, not theoretical knowledge.
3. Build a "Verification Habit" Into Every AI Interaction
Critical thinking is the dimension that separates Developing from Proficient. The simplest way to improve: before you use any AI output in your work, spend 60 seconds asking three questions:
- Is this factually accurate? Spot-check one specific claim.
- Is anything missing? AI tends to give confident, incomplete answers.
- Would I stake my reputation on this? If not, verify further.
This habit alone can shift your Critical Thinking score meaningfully. It's also the behaviour most visible to managers — they notice when you catch something an AI got wrong.
4. Document One AI Workflow Win Per Month
Promotion cases are built on evidence. Start keeping a simple log of times you used AI to produce better work, faster. Be specific:
- "Used Claude to draft the Q3 competitive analysis. First draft took 20 minutes instead of 3 hours. I verified all market share figures against primary sources and corrected two."
- "Built a prompt chain for weekly reporting that reduced my Friday synthesis from 90 minutes to 25 minutes."
These examples become talking points in your review. They also demonstrate Workflow & Application skills — the highest-weighted dimension. For ideas on what to include, see AI Skills for Your Resume: What to List, How to Prove.
5. Get the Certificate and Put It on LinkedIn
Once you've improved and retaken the assessment, your AISA score comes with a certificate you can add to your LinkedIn profile. This matters because it shifts the conversation from "I'm good at AI" (subjective) to "I scored X on a validated assessment" (verifiable). AISA results are stable — 98% of retakers land in the same or adjacent tier — so the certificate represents a durable signal, not a one-time lucky score.
The certificate is especially useful if you're in a role where AI skills aren't yet formally evaluated. It gives your manager something concrete to reference when making the case for your promotion to their leadership.
What Your Score Tells Your Manager
The real value of knowing your AI fluency score isn't the number itself — it's the conversation it enables. When you can say "I scored 58, here's what I'm doing to get to 65," you're demonstrating exactly the kind of self-awareness and growth orientation that promotion committees look for.
Consider the motivation data: professionals who take AISA for career transition purposes average 36.8, while those motivated by leadership average 51.4. The gap suggests that people already in leadership tracks have been building these skills longer. If you're aiming for that track, starting now puts you ahead of most of your peers.
The most important AI skills for employees aren't exotic. They're practical: clear communication with AI tools, critical evaluation of output, responsible data handling, and the ability to integrate AI into existing workflows. These are learnable skills with measurable benchmarks.
Your next review cycle is coming. The question isn't whether AI skills will factor into the decision — it's whether you'll know your score before your manager forms their own estimate.
Take the AISA assessment and find out where you actually stand. It takes one conversation, and you'll walk away with a score, a detailed breakdown, and a certificate — all of which are more useful than wondering.
Related reading: How Good Am I at AI? — The original self-assessment question, answered with data.
Related reading: What Is a Good AI Score? 2026 Benchmarks — Where the score thresholds come from and what they mean.
Related reading: AI Skills by Motivation: Why You Learn Predicts Your Score — How your reason for learning AI shapes your actual performance.
Frequently Asked Questions
Do AI skills affect promotions?
Yes. As AI adoption becomes standard across organisations, managers increasingly evaluate whether team members can use AI tools effectively as part of their overall performance. The World Economic Forum identified AI literacy as one of the fastest-growing required skills through 2030. While few companies have formal AI skill requirements for promotion yet, the ability to work productively with AI is becoming an implicit expectation — especially for mid-level and senior roles.
What AI score do I need for career advancement?
The Proficient tier (composite score of 60-79) represents the practical threshold where someone demonstrates consistent, effective AI use across work contexts. Across 2,578 AISA assessments, only 23.5% of people reach Proficient or above, with the overall average sitting at 46.2. Reaching Proficient puts you ahead of roughly three-quarters of assessed professionals and signals to managers that you can be trusted to use AI independently.
How do I prove AI skills to my manager?
The most effective approach combines an objective assessment score with documented examples of AI-driven work outcomes. Take a validated AI fluency assessment to establish your baseline, then maintain a log of specific instances where AI improved your output — including time saved, errors caught, and workflows redesigned. An AISA certificate provides a verifiable credential you can add to LinkedIn and reference in performance reviews, turning a subjective claim into concrete evidence.

Curious about your AI Fluency?
AISA helps you measure, prove and improve your AI skills — free report in a 20–40 minute chat.

