TL;DR: Healthy AI use is intentional, bounded, and supportive of your capabilities. Addiction is reflexive, expanding, and compensatory for capability loss. The distinction is behavioral, not about frequency.

The Short Version

You use AI every day. Does that mean you’re addicted? Not necessarily. Someone could use AI responsibly for 8 hours a day. Someone else could use it 30 minutes a day compulsively. The frequency doesn’t matter. The relationship does.

This framework cuts through the confusion: addiction vs. health is about intentionality, boundaries, capability impact, and control.

The Four Diagnostic Dimensions

Dimension 1: Intentionality vs. Reflex

Healthy Use:

  • You decide to use AI for a specific purpose
  • You open the tool with a clear task in mind
  • You close it when the task is complete
  • You could articulate why you’re using it before you use it

Addiction:

  • You open AI without a clear purpose
  • You check it reflexively (like checking your phone)
  • You keep it open to explore/tinker/see what comes up
  • You can’t easily articulate why you just opened it

This is the clearest diagnostic. Healthy users are intentional. Addicted users are reflexive.

In practice: Before you open your AI tool, can you finish this sentence? “I’m opening AI to [specific task].” If you can, you’re intentional. If you can’t, you’re reflexive.

💡 Key Insight: The difference between using a tool and being used by a tool is intentionality. Addiction begins when the tool starts making the decision about when to engage.

Dimension 2: Bounded vs. Expanding

Healthy Use:

  • You use AI for specific task categories (e.g., “boilerplate code” or “research synthesis”)
  • Your use cases have been stable for months
  • You have clear tasks you do without AI
  • You could list what you use AI for and what you don’t

Addiction:

  • Your use cases expand over time
  • This week you’re using it for X; next week you add Y and Z
  • You’re increasingly using it for everything
  • You notice yourself reaching for it for tasks you used to do alone
  • You can’t easily list what you do without AI

Healthy use is bounded. You’ve decided: “AI helps with these tasks, not those.” Addiction is expanding. The boundary keeps moving outward.

In practice: Look at your AI use from 3 months ago vs. now. Has it expanded? Are you using it for more types of tasks? If yes, you’re in the expansion phase. This often precedes full-scale addiction.

Dimension 3: Capability Support vs. Capability Replacement

Healthy Use:

  • AI augments your capabilities
  • You’re still capable without it (just slower)
  • You understand how to do the work yourself
  • AI helps you do better work, not different work
  • Your skills in the relevant domains are stable or improving

Addiction:

  • AI replaces your capabilities
  • You’re not capable without it (you’d be stuck)
  • You don’t remember how to do the work yourself
  • AI is doing work you can’t do anymore
  • Your skills in relevant domains are declining

This is the capability test. Does AI support what you can do, or does it compensate for what you’ve stopped being able to do?

In practice: For a task you use AI for, could you do it yourself if you had to? Not in the optimal way, but could you do it? If yes, AI is supporting. If no, it’s replacing.

📊 Data Point: Capability loss is measurable after 4-6 weeks of non-use for cognitive tasks; if you haven’t done a task without AI in that timeframe, you’ve likely lost the skill.

Dimension 4: Internal Locus vs. External Locus

Healthy Use:

  • You decide when to use AI
  • You could stop using it and adjust your workflow
  • You experience no anxiety at access loss
  • You feel in control of your tool use

Addiction:

  • AI use decides you (you feel compelled)
  • Stopping would feel impossible or create anxiety
  • You experience anxiety/stress when the tool is unavailable
  • You feel the tool is controlling your behavior

This is the autonomy test. Do you control the tool, or does it control you?

In practice: Imagine your AI tools were unavailable for a week. What’s your genuine internal response? Neutral adjustment (healthy) or anxiety/stress (addicted)?

The Integrated Framework

You can score yourself on each dimension:

For each dimension, rate yourself:

  • Mostly healthy = 0 points
  • Mixed = 1 point
  • Mostly addicted = 2 points

Total score:

  • 0-2: Healthy use. You might have minor boundary issues, but you’re largely in control.
  • 3-5: Mixed/at-risk. You have healthy use in some areas and addiction-like patterns in others. This is where intervention helps most.
  • 6-8: Addiction-range. Multiple dimensions indicate dependency. Changes needed.

Importantly: This isn’t about judgment. It’s about accurate diagnosis. You need to know where you are to decide where you want to be.

The Progression: How Health Becomes Addiction

Most people don’t start addicted. Addiction develops through stages.

Stage 1: Tool Introduction (Weeks 1-2)

  • All four dimensions are healthy
  • You’re exploring a new tool
  • Intentional, bounded, supportive, autonomous

Stage 2: Optimization (Weeks 3-8)

  • Intentionality remains; you’re deliberately using the tool
  • Expansion begins; you find more use cases
  • Capability support is present; AI is augmenting
  • Autonomy intact; you choose when to use it

Stage 3: Integration (Weeks 9-16)

  • Intentionality decreases; you start reaching for it reflexively
  • Expansion accelerates; it’s becoming your default
  • Capability support transitions; you’re losing ability to work without it
  • Autonomy begins eroding; you feel you need it

Stage 4: Dependency (Month 4+)

  • Intentionality is gone; it’s reflexive
  • Expansion is complete; you use it for nearly everything
  • Capability replacement has occurred; you can’t work without it
  • Autonomy is lost; you feel compelled

This progression is normal for addictive tools. The question: at which stage do you intervene?

Most people don’t notice the progression until stage 4. By then, change is much harder. Intervention at stage 3 (when you notice expansion and first signs of capability loss) is much more effective.

The Nuance: Performance vs. Process

One complication: sometimes high AI use does improve performance.

A builder using AI 8 hours a day might ship more features than someone using it 2 hours a day. The metrics look good. Performance is actually improved. But are they addicted?

It depends on the other dimensions. If they’re:

  • Using it reflexively (not intentionally)
  • Expanding their use (not bounded)
  • Losing capability (not supporting)
  • Anxious without it (not autonomous)

Then yes, they’re addicted. The good performance is masking the dependency.

This is the hidden trap: addiction that works. You’re performing well, so you rationalize the whole thing. But the underlying dynamics are still addictive.

💡 Key Insight: Performance can improve while addiction develops. High output masks low autonomy. The mask prevents intervention until the bottom falls out.

The Honest Self-Assessment

To use this framework honestly:

  1. Score yourself on each dimension. Don’t strategically score. Actually assess: are you being intentional? Is your use bounded? Is AI supporting or replacing? Do you feel autonomous?

  2. Notice the pattern. Are multiple dimensions in addiction-range, or just one? One dimension in trouble is different from all four.

  3. Identify the progression stage. Where are you? Stage 1 (healthy), Stage 2-3 (at-risk), or Stage 4 (addicted)?

  4. Consider the delta. How has each dimension changed over the past 3 months? Improving or degrading?

  5. Make a decision. Do you want to change, or are you comfortable with where things are? Both are valid. Just be honest about the choice.

What This Means For You

Use this framework not as judgment, but as honest self-assessment. You probably have some dimensions in healthy range and some in addiction range. That’s normal.

The goal isn’t perfection. It’s awareness and intentional choice.

If you discover you’re in addiction-range on multiple dimensions, you have tools:

  • Refocus on intentionality: decide before you open the tool
  • Restore boundaries: pick use cases you allow, use cases you don’t
  • Support capability: practice non-AI versions of tasks
  • Rebuild autonomy: create friction, add accountability

You’ve already read the recovery patterns. This framework tells you which ones matter most for you.

Key Takeaways

  • Healthy use is intentional, bounded, capability-supporting, and autonomous
  • Addiction is reflexive, expanding, capability-replacing, and compulsive
  • Frequency of use is not diagnostic; relationship with use is
  • Progression from health to addiction happens in stages; intervention is easier earlier
  • High performance can mask addiction; good metrics don’t prove health
  • Self-assessment framework: score intentionality, boundedness, capability impact, and autonomy

Frequently Asked Questions

Q: I use AI a lot but I score as healthy. Does that mean I’m fine? A: Probably. Frequency isn’t diagnostic. If you’re intentional, bounded, still capable without it, and feel autonomous, high use can be fine.

Q: I use AI rarely but I score as addicted. How is that possible? A: It happens when your limited use is reflexive, you feel compelled, and you’re losing capability in those areas. Less frequent addiction still looks like addiction.

Q: Can someone be partially addicted? Addicted to AI for some tasks but not others? A: Yes. You might be addicted to AI for decision-making but healthy with it for boilerplate work. The framework helps you identify which areas need attention.

Not medical advice. Community-driven initiative. Related: Are You Dependent on AI? A 10-Question Self-Assessment | Signs You Are Addicted to AI | What AI Addiction Recovery Actually Looks Like