# Should I use ChatGPT for advice on getting my ex back?

> By Benny Lichtenwalner, breakup coach and author of "How to Get Your Ex Back".
> Canonical: https://bennylichtenwalner.com/answers/should-i-use-chatgpt-for-advice-on-getting-my-ex-back/

**Short answer:** Use it to analyze, never to decide. ChatGPT can help you understand an attachment style, interpret a behavioral pattern, or organize your thinking. The moment you ask it what to text, whether to reach out, or how to respond, you're outsourcing a judgment call to a tool with no stake in the outcome and no way to read the dynamic.

# Should I Use ChatGPT for Advice on Getting My Ex Back?

Use it to understand. Don't use it to decide.

That's the whole answer. But since most people can't hold that line when they're anxious, it's worth being specific about where it breaks.

Benny Lichtenwalner, breakup coach and author of *How to Get Your Ex Back*, has seen this pattern consistently: a client goes down an AI rabbit hole looking for action advice and comes back with something they shouldn't do. Not because the AI is malicious. Because it's generating plausible output from a summary of the situation, and a summary of your situation is not your situation.

## What AI Tools Are and Aren't Good For Here

| Use case | Useful | Not useful |
|---|---|---|
| Understanding attachment styles conceptually | Yes | — |
| Interpreting a behavioral pattern you've observed | Carefully | — |
| Organizing your own thinking | Yes | — |
| Deciding what to text your ex | — | No |
| Figuring out whether to reach out | — | No |
| Getting a second opinion on your tactical plan | — | No |

The line is clear: analysis of what you're observing is one thing. Action advice — what to say, whether to move, how to respond — is something else. The second category requires reading a dynamic that an AI, working from your compressed description of it, cannot actually read.

## Why Does This Feel Like Research When It Isn't?

You're not gathering information when you ask an AI what to do. You're looking for permission — permission to act, or permission to wait. You want certainty, and you're hoping an authoritative-sounding response will provide it.

It won't. A tool that generates confident answers from incomplete data isn't reducing your uncertainty. It's covering it with language.

You're also trying to solve a math problem before the teacher has told you what the variables stand for. The plan accounts for the ambiguity you're sitting with. Your anxiety does not. Running to an AI for action advice is what happens when anxiety wins the argument.

## What to Do Instead

1. **Write down what you actually observed.** Not what you think it means. What did your ex say or do, in concrete terms?
2. **Separate observation from interpretation.** Hold those two things apart deliberately before doing anything else.
3. **Continue the plan.** The plan was built for exactly this kind of ambiguity. Trust it.
4. **Use AI for conceptual research, not tactical decisions.** Understanding what a dismissive-avoidant does under stress is useful. Asking what to text a dismissive-avoidant who sent you a specific message is not.
5. **Assess each interaction fresh.** Each exchange is a new data point. Don't pre-load conclusions based on AI output from two days ago.

## What NOT to Do

Don't paste your ex's messages into an AI tool and ask what they mean. The interpretation you get back will be shaped by how you framed the situation, which was shaped by what you already feared. That's confirmation bias with extra steps.

Don't ask AI to draft your next message to your ex. Output optimized to sound reasonable is not the same as output calibrated to your specific dynamic.

Don't treat an AI response as permission to break the plan. "It said I should reach out" is not a reason to reach out.

Don't convince yourself you're the exception to how your ex's attachment style behaves because an AI generated a scenario where things work out differently. That's a version of overthinking too.

## When It Goes Wrong

Two clients came to me after going down AI rabbit holes. Both were in no contact, both had ambiguous signals from their ex, and both had asked an AI whether they should reach out.

One got an AI response saying their ex was probably withdrawing and might respond to a light, friendly check-in. They sent it. Their ex replied briefly and went quiet again. Three weeks of work reset.

The other got nearly identical AI advice and brought it to me first. I said wait. They waited. Their ex reached out four days later.

The AI advice wasn't wrong exactly. It generated a plausible-sounding response. What it couldn't do was read the actual dynamic well enough to know that waiting was right and acting was wrong. That's the gap no prompt can close.

Use it to understand. Trust the plan to decide.