Individual AI Self-Assessment

A personal AI maturity self-assessment covering strategy, skills, workflow integration, governance, and impact — helping individuals identify strengths and growth areas across nine dimensions.

Run in TeamRetro

How well are you using AI — really?

Effective AI adoption starts with the individual. This self-assessment helps team members honestly evaluate where they stand across nine dimensions of personal AI maturity — from strategic alignment and tool adoption through to governance, output quality, and continuous improvement. Each dimension follows a five-stage progression from Ad Hoc to Optimized, giving individuals a clear picture of their current practice and a concrete path forward. Use this assessment to spark meaningful conversations about AI skills, identify personal development priorities, and build a shared language for what good AI use looks like at the individual level. Whether you're just starting out or already embedding AI deeply into your daily work, this health check surfaces the gaps and strengths that matter most.

Dimensions

AI Strategy & Leadership

How well your AI use is aligned to organisational goals, prioritised thoughtfully, and owned with accountability.

  • Strategic Alignment

    I use AI in ways that clearly support my team, department, or organisational goals.

    1. Ad HocAI use is opportunistic; little connection to broader goals.
    2. EmergingSome AI use ties back to goals; much of it doesn't.
    3. DefinedAI use is anchored to goals that matter to the team or organisation.
    4. ManagedAI use is prioritised against goals; lower-value uses are dropped deliberately.
    5. OptimizedAI is a deliberate lever for the outcomes that matter most.

  • Use Case Prioritisation

    I choose where to apply AI based on value, feasibility, and risk — not on novelty.

    1. Ad HocAI is applied wherever it looks interesting; trade-offs aren't considered.
    2. EmergingSome use cases are chosen thoughtfully; many are not.
    3. DefinedUse-case decisions weigh value, feasibility, and risk.
    4. ManagedPrioritisation is consistent and reviewed; weaker use cases are dropped.
    5. OptimizedAI is applied where it produces the most return for the effort, not where it's shiny.

  • Leadership Awareness

    My manager and leadership know how I use AI and have given clear expectations about it.

    1. Ad HocLeadership doesn't know how I use AI; I don't know what is expected.
    2. EmergingSome informal awareness exists; expectations are unclear.
    3. DefinedLeadership knows broadly how I use AI; expectations are stated.
    4. ManagedAI use is part of normal conversations with my manager; expectations are clear and current.
    5. OptimizedAI use is openly discussed and supported by leadership; I operate with confidence about what is expected.

  • Personal Ownership

    I take responsibility for what AI does on my behalf — the outputs, the decisions, and the consequences.

    1. Ad HocWhen AI gets something wrong, the blame goes to the tool; ownership is unclear.
    2. EmergingOwnership of AI outputs is mixed; sometimes mine, sometimes the tool's.
    3. DefinedI own what I send out, regardless of whether AI helped produce it.
    4. ManagedOwnership of AI outputs is automatic; mistakes are mine to learn from.
    5. OptimizedPersonal accountability for AI-assisted work is total; no hiding behind the tool.

Tool Adoption & Access

Whether you have the right AI tools, use them consistently, and stay current on new capabilities relevant to your work.

  • Tool Availability

    I have access to the AI tools I need to do my work well.

    1. Ad HocAccess to AI tools is patchy; I work around gaps.
    2. EmergingSome AI tools are available; others I'd need are not.
    3. DefinedThe AI tools I need to do my work are available to me.
    4. ManagedAccess keeps pace with my work; new tools arrive when they're justified.
    5. OptimizedAI tooling is a strength of my setup; nothing useful is out of reach.

  • Tool Fit

    The AI tools I use fit the tasks I actually do.

    1. Ad HocI use whatever AI tool I happened to start with; fit is incidental.
    2. EmergingSome tools fit my work well; others feel like a stretch.
    3. DefinedMy AI tools match the work I do most often.
    4. ManagedTool choices are reviewed against actual tasks; misfits are swapped out.
    5. OptimizedTool fit is excellent; AI tools feel built for the work I do.

  • Personal Adoption Consistency

    I use AI consistently in the places it adds value, rather than reaching for it only occasionally.

    1. Ad HocAI use is sporadic; I forget it's an option for many tasks.
    2. EmergingAI is part of some tasks; consistency is uneven.
    3. DefinedAI use is consistent for the tasks where it adds value.
    4. ManagedAI use is habitual; I recognise when to use it without thinking.
    5. OptimizedAI is so embedded in my work that the question of whether to use it no longer comes up.

  • Awareness of New Capabilities

    I keep myself current on what AI can newly do that is relevant to my work.

    1. Ad HocI hear about new AI capabilities by accident, long after they appear.
    2. EmergingSome awareness; I lag behind the state of the art.
    3. DefinedI follow AI developments relevant to my work and try new capabilities deliberately.
    4. ManagedKeeping current is a habit; I evaluate and adopt new capabilities on a regular cadence.
    5. OptimizedI'm on the leading edge for my role; new capabilities reach my work fast and considered.

AI Skills & Confidence

Your ability to understand AI, craft effective prompts, think critically about outputs, and grow your skills over time.

  • AI Literacy

    I understand what AI can and cannot do, and where its limits sit.

    1. Ad HocMy mental model of AI is fuzzy; I'm often surprised by what it can or can't do.
    2. EmergingSome grasp of capabilities; gaps and limits are unclear.
    3. DefinedI have a working understanding of what AI does well, where it struggles, and why.
    4. ManagedMy mental model is current and accurate; I update it as AI evolves.
    5. OptimizedI can predict AI behaviour and limits with confidence; my mental model is a working tool.

  • Prompting Skill

    I give AI clear instructions, context, and constraints to get useful output.

    1. Ad HocPrompts are one-line questions; outputs reflect that.
    2. EmergingSome prompts are richer; results are inconsistent.
    3. DefinedPrompts include context, constraints, and examples where useful.
    4. ManagedPrompting is a deliberate craft; I tune prompts and reuse what works.
    5. OptimizedPrompting is a strong personal skill; I get high-quality output with confidence.

  • Critical Thinking

    I question, check, and refine AI-generated outputs rather than taking them at face value.

    1. Ad HocAI output is treated as correct unless something obvious is wrong.
    2. EmergingSome checking happens; depends on the day.
    3. DefinedAI output is critically reviewed before I use it.
    4. ManagedCritical review is automatic; I notice subtle errors and refine outputs deliberately.
    5. OptimizedI'm a calibrated sceptic of AI output; trust is earned, not assumed.

  • Personal Learning & Development

    I invest deliberately in growing my AI skills.

    1. Ad HocSkill growth is whatever I pick up by accident.
    2. EmergingSome learning happens; it's not deliberate.
    3. DefinedI set aside time to learn AI skills relevant to my work.
    4. ManagedLearning is part of how I work; I notice gaps and close them.
    5. OptimizedSkill development on AI is continuous and intentional; my capability compounds over time.

Workflow Integration

How deeply AI is embedded in your daily work, standard processes, and the balance between AI and human judgment.

  • Daily Use

    AI is naturally part of my everyday work.

    1. Ad HocMost days pass without using AI; it's an occasional tool.
    2. EmergingAI shows up in some daily work; not most.
    3. DefinedAI is part of my daily work for the tasks where it helps.
    4. ManagedAI is a default reach; not using it is the exception.
    5. OptimizedAI is woven into how I work; the seam is invisible.

  • Process Integration

    AI is built into my standard ways of working, not bolted on at the edges.

    1. Ad HocAI sits alongside my workflow; I copy and paste in and out.
    2. EmergingSome processes have AI integrated; many do not.
    3. DefinedAI is part of my standard processes for the work where it adds value.
    4. ManagedAI integration is reviewed and improved; new processes are designed with AI in mind.
    5. OptimizedMy processes are AI-native where it makes sense; the workflow assumes AI is part of it.

  • Human–AI Balance

    I know when to rely on AI and when my own judgment, expertise, or review is essential.

    1. Ad HocThe boundary between AI and my own judgment is unclear; I lean too far one way or the other.
    2. EmergingSome sense of when to trust AI; rules are ad hoc.
    3. DefinedI have a working sense of when AI is enough and when my judgment is required.
    4. ManagedThe balance is deliberate and reviewed; I shift it as my experience grows.
    5. OptimizedThe balance is intuitive and correct; AI extends my judgment without replacing it.

  • Work Redesign

    I rethink how I work to take advantage of what AI changes — not just add AI to old habits.

    1. Ad HocMy habits are unchanged; AI is bolted on to old ways.
    2. EmergingSome habits have shifted; most remain the same.
    3. DefinedI've redesigned parts of my work to take advantage of AI.
    4. ManagedWork redesign is ongoing; I question old habits as AI capabilities grow.
    5. OptimizedMy way of working is fundamentally shaped by AI; nothing is on autopilot from the pre-AI era.

Data, Knowledge & Context

The quality of information you bring to AI, your access to internal knowledge, and how responsibly you handle data.

  • Data Quality

    The information I use with AI tools is reliable, accurate, and up to date.

    1. Ad HocI feed AI whatever I can find; data quality is incidental.
    2. EmergingSome attention to data quality; gaps are common.
    3. DefinedThe data I give AI is fit for the question I'm asking.
    4. ManagedI check and curate data before AI uses it; poor inputs are caught.
    5. OptimizedData quality is a habit; AI gets the best information I have, every time.

  • Knowledge Access

    I can easily get to the internal knowledge I need for AI-assisted work.

    1. Ad HocFinding internal knowledge to feed AI is slow and uncertain.
    2. EmergingSome sources are easy to reach; others are buried.
    3. DefinedI can reach the internal knowledge I need for AI work without friction.
    4. ManagedAccess is fast and reliable; I notice and flag gaps.
    5. OptimizedKnowledge access is a non-issue; AI-assisted work is unblocked by it.

  • Context Provision

    I give AI the right background, examples, and constraints so its output fits my situation.

    1. Ad HocPrompts are bare; AI lacks the context it needs and shows it.
    2. EmergingSome prompts have context; depth and quality are uneven.
    3. DefinedPrompts include the context AI needs to be useful.
    4. ManagedContext provision is a habit; AI rarely produces output that misses the situation.
    5. OptimizedContext is provided so well that AI output lands close to what I need on first pass.

  • Personal Data Stewardship

    I manage the data and knowledge I work with responsibly — accuracy, ownership, and appropriate use.

    1. Ad HocData stewardship is not something I think about; AI gets whatever I have.
    2. EmergingSome attention to stewardship; it's reactive.
    3. DefinedI treat the data I use with AI as something I'm responsible for.
    4. ManagedStewardship is part of how I work; I check ownership and appropriate use before sharing data with AI.
    5. OptimizedData stewardship is automatic; I can defend the appropriateness of every piece of data I give AI.

Governance, Risk & Compliance

Your awareness of AI policy, privacy obligations, bias risks, and your ability to account for how AI was used.

  • Policy Awareness

    I know my organisation's AI policies, boundaries, and approval requirements.

    1. Ad HocI don't know what the AI policy is; I act on assumption.
    2. EmergingSome awareness of policy; details are hazy.
    3. DefinedI know the AI policy and operate within it.
    4. ManagedPolicy awareness is current; I track changes and adapt.
    5. OptimizedPolicy is part of how I think about AI use; I can explain it to others confidently.

  • Privacy & Confidentiality

    I keep sensitive, personal, or restricted information out of AI tools that shouldn't see it.

    1. Ad HocI paste in whatever I'm working with; privacy isn't front of mind.
    2. EmergingSome attention to what goes in; mistakes happen.
    3. DefinedI check before sharing sensitive information with AI; rules are clear in my head.
    4. ManagedPrivacy is automatic; sensitive data doesn't reach AI it shouldn't.
    5. OptimizedPrivacy and confidentiality are a settled discipline; I'd notice immediately if a boundary was at risk.

  • Bias & Fairness

    I think about bias, fairness, and potential harm in AI-assisted output before using it.

    1. Ad HocBias and fairness aren't something I consider when using AI.
    2. EmergingSome awareness; checking is rare.
    3. DefinedI review AI output for bias and fairness when the stakes warrant it.
    4. ManagedBias awareness is part of how I review AI output; I've caught and corrected issues.
    5. OptimizedFairness is a default lens on AI output; I can defend my use of AI on this dimension.

  • Auditability

    I keep enough of a trail to explain when, how, and why I used AI for work that matters.

    1. Ad HocNo trail; if asked how AI was used, I'd struggle to reconstruct it.
    2. EmergingSome trail exists, mostly in my head.
    3. DefinedFor work that matters, I can explain how AI contributed.
    4. ManagedAI use is documented where it counts; reviewers and managers can follow the path.
    5. OptimizedAI use is auditable end-to-end for important work; nothing is opaque.

Output Quality & Human Review

How rigorously you verify AI-generated content, maintain your quality standards, and learn from AI errors.

  • Accuracy Checking

    I verify AI-generated information before relying on it.

    1. Ad HocAI output is used as-is; accuracy checks are rare.
    2. EmergingSome checking happens; depth depends on the day.
    3. DefinedAI output is verified before I rely on it for anything that matters.
    4. ManagedVerification is a habit; errors are caught early.
    5. OptimizedAccuracy checking is automatic and calibrated; I notice subtle wrongness fast.

  • Personal Quality Standards

    My AI-assisted output meets my standards for clarity, accuracy, and professionalism.

    1. Ad HocAI output goes out lightly edited; quality is uneven.
    2. EmergingSome output is polished; some isn't.
    3. DefinedAI-assisted output meets my standards before it leaves me.
    4. ManagedQuality is consistent and improving; AI extends the bar I can hold.
    5. OptimizedAI-assisted output is indistinguishable from my best work; quality is a settled property.

  • Human Oversight

    I have a clear sense of when human review or approval is required for AI-assisted work.

    1. Ad HocWhether AI work needs human review is decided case by case; rules are unclear.
    2. EmergingSome review thresholds exist; coverage is patchy.
    3. DefinedI know which AI-assisted work needs human review and route it accordingly.
    4. ManagedReview thresholds are deliberate and refined over time.
    5. OptimizedOversight is automatic and proportionate; high-stakes AI work gets review, low-stakes doesn't.

  • Error Learning

    I learn from AI mistakes, hallucinations, and poor outputs rather than just moving on.

    1. Ad HocWhen AI gets things wrong, I move on; nothing changes.
    2. EmergingSome lessons are absorbed; many aren't.
    3. DefinedAI mistakes inform how I prompt, check, and use AI next time.
    4. ManagedErrors compound into a better personal practice over time.
    5. OptimizedAI errors are short-lived; I rarely repeat the same mistake.

Collaboration & Knowledge Sharing

How actively you share AI knowledge, experiment with new approaches, learn from others, and build reusable assets.

  • Sharing What Works

    I share useful prompts, examples, workflows, and lessons with my team rather than hoarding them.

    1. Ad HocSharing rarely happens; what works stays with me.
    2. EmergingSome sharing; it depends on whether I remember.
    3. DefinedI share useful AI work with the team as a normal part of how I operate.
    4. ManagedSharing produces lift for the team; my contributions are recognised and used.
    5. OptimizedSharing is automatic; my team is better at AI because of my contributions.

  • Personal Experimentation

    I feel safe trying new things with AI and learning from what doesn't work.

    1. Ad HocExperimentation feels risky; I stick to what I know.
    2. EmergingSome experimentation happens; I'm cautious about it.
    3. DefinedI experiment with AI deliberately and learn from what fails.
    4. ManagedExperimentation is a habit; failure is treated as part of learning.
    5. OptimizedI'm fearlessly experimental with AI; new capabilities reach my work fast because of it.

  • Learning From Others

    I learn from how other people, teams, and external examples use AI well.

    1. Ad HocI work AI out alone; little input from others.
    2. EmergingSome learning from others; it's informal.
    3. DefinedI deliberately learn from how others use AI well.
    4. ManagedLearning from others is a habit; I bring patterns into my own work.
    5. OptimizedMy AI practice is shaped by what works elsewhere; I import and adapt fast.

  • Reusable Personal Assets

    I build reusable AI templates, prompt libraries, checklists, or playbooks for the work I do often.

    1. Ad HocEvery AI task starts from scratch; nothing is reused.
    2. EmergingSome prompts and templates are kept; reuse is limited.
    3. DefinedI have a working set of reusable AI assets for my common tasks.
    4. ManagedReusable assets are maintained and improved; they save time visibly.
    5. OptimizedMy reusable AI assets are a real personal advantage; they compound my output.

Impact Measurement & Improvement

Whether you track the real impact of your AI use on productivity and quality, and continuously improve your practice.

  • Personal Productivity Impact

    I notice whether AI is actually saving me time or increasing my capacity.

    1. Ad HocWhether AI saves me time is unexamined; I assume it does.
    2. EmergingSome sense of impact; mostly anecdotal.
    3. DefinedI notice where AI saves time and where it doesn't.
    4. ManagedProductivity impact is monitored; low-value AI use is dropped.
    5. OptimizedAI's contribution to my productivity is clear and acted on; effort goes where return is real.

  • Quality Impact

    I notice whether AI improves the quality, consistency, or usefulness of my work.

    1. Ad HocQuality impact of AI is unexamined.
    2. EmergingSome sense of quality lift; uneven.
    3. DefinedI notice where AI lifts quality and where it doesn't.
    4. ManagedQuality impact informs where I use AI and where I stop.
    5. OptimizedAI's quality contribution is clear; it's used where it lifts work and avoided where it doesn't.

  • Outcome Connection

    I can connect my AI use to outcomes that matter — customer value, time saved, mistakes avoided, work shipped.

    1. Ad HocAI use isn't connected to outcomes I can name.
    2. EmergingSome outcomes are visible; the connection is partial.
    3. DefinedI can point to outcomes that AI changed for the better.
    4. ManagedOutcomes drive how I deploy AI; I shift effort to where it counts.
    5. OptimizedAI is connected to outcomes precisely; I can name the difference it makes.

  • Continuous Improvement of My AI Practice

    I use feedback, results, and lessons to keep improving how I use AI.

    1. Ad HocHow I use AI doesn't change; first habits stick.
    2. EmergingOccasional improvements; nothing systematic.
    3. DefinedI revise how I use AI based on what works and what doesn't.
    4. ManagedImprovement is a real loop — try, observe, adjust — and the changes hold.
    5. OptimizedMy AI practice gets sharper continuously; nothing about how I use AI is static.

When to use this health check

  • When individuals want an honest baseline of their current AI practice before starting a development plan.
  • During team AI capability reviews to surface variation in skills, habits, and governance awareness.
  • As part of an onboarding process to help new team members understand the expected standard of AI use.
  • When an organisation is rolling out new AI tools or policies and wants to assess readiness.
  • Periodically (e.g. quarterly) to track personal AI maturity growth over time.
  • When a team wants to identify shared gaps and prioritise collective AI learning investments.

Tips & tricks

  • Encourage honest self-rating — this is a personal development tool, not a performance review. Psychological safety matters.
  • Run the assessment individually first, then share results as a team to identify common themes and outliers.
  • Focus discussion on the dimensions with the widest spread across the team — these often reveal the biggest opportunities.
  • Pair the assessment with a concrete action: each person picks one dimension to improve before the next check-in.
  • Use the Ad Hoc and Optimized descriptions as conversation starters — ask 'what would Optimized look like for us specifically?'
  • Revisit the assessment every quarter to track progress and celebrate movement up the maturity scale.
  • The 'Not Applicable' option is there for roles where certain dimensions genuinely don't apply — don't force a rating.

Frequently asked questions

Who should complete this self-assessment?
Anyone who uses AI tools as part of their work — from occasional users to power users. It's designed for individuals across all roles and functions, not just technical teams.
How long does the assessment take?
Most people complete it in 10–15 minutes. The matrix format lets you move through all nine dimension groups efficiently.
Should I share my results with my manager?
That's up to you and your team culture. The assessment is most valuable as a personal development tool, but sharing results with your manager or team can unlock coaching conversations and collective improvement.
What do the five maturity levels mean?
The levels follow a standard progression: Ad Hoc (unstructured, reactive), Emerging (some awareness but inconsistent), Defined (clear and consistent practice), Managed (deliberate and reviewed), and Optimized (continuous improvement, leading practice).
How often should I run this assessment?
Quarterly is a good cadence for most teams. It gives enough time for meaningful change while keeping the feedback loop tight enough to stay relevant.
What if a dimension doesn't apply to my role?
Use the 'Not Applicable' option. Not every dimension will be relevant to every role — forcing a rating where it doesn't apply skews the results.
Can this be used as a team-level assessment rather than individual?
Yes. Running it individually and then aggregating results gives a team-level view of AI maturity, which is useful for identifying shared gaps and planning collective learning.