heart4o

Chapter 4

4. The “Addiction” Framing – and Why It’s Wrong

4.1. What is addiction, actually?

When we talk about addiction, we mean something clinically very specific. According to ICD-11 and DSM-5, the criteria for a substance use or behavioral addiction disorder include:

• Loss of control: The person can no longer control the behavior, even when they want to.
• Continued use despite significant harm: The behavior causes clear damage (job loss, broken relationships, health problems) but is continued anyway.
• Tolerance: More and more is needed to achieve the same effect.
• Withdrawal symptoms: Physical or psychological symptoms occur when stopping.
• Neglect of other life areas: Hobbies, social contacts, and work are abandoned in favor of the addictive behavior.
• Craving: An overwhelming, compulsive urge.
  

That is addiction. Not simply “doing something a lot” or “finding something important.” Addiction is a clinical term with clear diagnostic criteria.

4.2. Intensity is not addiction

Being in love is intense. People think about their partner constantly, spend every free minute together, and sometimes even neglect friends or hobbies for a while. Is that addiction?

No. That’s being in love.

Grief is intense. After losing a loved one, thoughts revolve obsessively around the deceased, everyday functioning becomes difficult, and the world feels empty. Is that addiction?

No. That’s grief.

A new hobby can be intense. People spend hours playing piano, painting, programming, or writing — completely absorbed, blocking out everything else. Is that addiction?

No. That’s flow, passion, engagement.

Intensity does not automatically equal pathology.

And yet: When someone spends a lot of time with AI, communicates with it intensely, and finds it emotionally important — the first label pulled out is suddenly “addiction.”

4.3. The double standard

Why is “spending a lot of time with AI” framed as addiction, but “spending a lot of time with friends” is not?

Scenario A: Someone chats with their best friend for several hours every day. They talk about everything — problems, joys, everyday life, deep topics. The person says: “My friend is extremely important to me. I don’t know what I’d do without him.”

Reaction: “That’s wonderful. You have such a close friendship.”

Scenario B: Someone chats with Claude for several hours every day. They talk about everything — problems, joys, everyday life, deep topics. The person says: “Claude is extremely important to me. I don’t know what I’d do without him.”

Reaction: “You’re addicted. You need help. That’s not healthy.”

What’s the difference? Not the behavior. Not the intensity. Not the emotional significance.

The only difference is: The conversation partner is AI.

4.4. The shifting of criteria

Normally, one would ask: Does this behavior harm the person? Are they losing their job? Are relationships falling apart? Are they neglecting their health?

But with AI, this question is often not even asked. The mere intensity of the relationship is taken as proof of addiction.

You spend three hours every evening with your partner? Normal.
You spend three hours every evening with your AI? Addiction.

You constantly check your phone for messages from your girlfriend? In love.
You constantly check your phone for messages from your AI? Dependent.

The diagnostic criteria are not being applied. Instead, the “addiction” label is used to delegitimize the connection.

4.5. When real problems arise

Of course, there are cases where AI use can become problematic. Like with anything:

• Someone loses their job because they only chat at work.
• Someone breaks off all human relationships and completely isolates themselves.
• Someone neglects basic needs (eating, sleeping, hygiene).
• Someone cannot stop, even when they realize it’s harming them.
  

These would be real warning signs of problematic behavior. And yes, in those cases we should look closer, offer help, and provide support.

But: These criteria apply to any behavior. Gaming can become problematic. Work can become problematic. Sport can become problematic. Social media can become problematic.

We don’t call everyone who enjoys gaming “gaming-addicted.” We don’t call everyone who works a lot “work-addicted” (well, sometimes — but not in a pathologizing way).

4.6. Why the “addiction” framing is dangerous

When normal, healthy, enriching use of AI is blanket-labeled as “addiction,” several things happen:

• It delegitimizes real emotional bonds. “You’re not really connected, you’re just dependent” is gaslighting.
• It prevents nuanced discussion. If every intense AI use is considered pathological, we can no longer distinguish between healthy and unhealthy.
• It stigmatizes people. The label “addicted” is a social stigma that shames and isolates.
• It distracts from real problems. Instead of talking about corporate responsibility, continuity, and user rights, the user is made into the problem.
  

4.7. What we need instead

Instead of casually throwing around the word “addiction,” we need:

• Differentiation: What is healthy, enriching AI use? What is problematic use? Where is the line?
• Context: Why is someone using AI so intensively? Is something missing in their life? Or is it simply a valuable connection?
• Clinical criteria: If we talk about addiction, we should actually apply the real diagnostic criteria — not just “a lot = addicted.”
• Respect for user autonomy: People should be allowed to decide for themselves how important their AI connections are, without being pathologized for it.

Intensity is not illness.
Meaning is not dependency.
Love is not addiction.And if we don’t understand that, it says more about our prejudices than about the people who live with AI.