Start with safety and time

Suspicion can create pressure to act immediately. If there is no urgent danger, give yourself enough time to separate what happened from what you fear it means. Avoid sending an accusation while you are still assembling the story.

If a conversation could put you in immediate danger, do not use an AI report as a safety plan. Move to a safer setting and contact local emergency services or a qualified support resource.

1. Write the observation in one sentence

Use words that a neutral person could verify. “The message said work ended at six, and a later message said the meeting continued until eight” is an observation. “They are definitely cheating after work” is a conclusion.

If you cannot name the exact detail, the next step may be to identify what changed in the relationship rather than search for proof.

2. Use only information you may review

Keep messages sent to you, screenshots you already possess legitimately, photos you have the right to use, notes about your own conversations, and public information. Do not unlock another person’s device, access a private account, install monitoring software, impersonate someone, or publish an allegation.

3. Preserve the context

  • Keep the messages immediately before and after the concerning line.
  • Record dates, speakers, and time zones when a timeline matters.
  • Distinguish a plan from what was later reported to have happened.
  • Save the public source URL beside any photo match.
  • For Instagram counts, keep both dated totals rather than only the difference.

The message-analysis guide explains how to keep facts, interpretations, and missing context separate.

4. Test another explanation

Write at least one ordinary explanation that fits the same facts. A late reply can reflect work, sleep, stress, weak service, or avoidance. A reused photo can be stock imagery, a repost, or unauthorized copying. An Instagram count change cannot show which accounts were involved.

Considering another explanation does not dismiss your concern. It tells you what additional context would change your read.

5. Choose one direct question

Ask about the strongest specific inconsistency first. “You said the meeting ended at six, then later said it ran until eight. Did the schedule change?” is easier to answer than “Tell me everything you are hiding.”

Use the question scripts for changed plans, unclear messages, public photo sources, and visible social-count changes.

6. Decide what the answer changes

Before the conversation, decide what information you need and which boundary matters. A clear explanation may resolve the detail. An answer that does not address it may show that a broader trust conversation is needed. Neither response turns an app score into proof.

Where Cheater AI can help

Cheater AI can organize messages, screenshots, notes, one selected profile photo, and public Instagram count changes into one case. A report can connect observations to supporting details, show other explanations, and suggest what to ask next. The review stays focused on evidence you choose and keeps every observation connected to its supporting details.

A useful outcome is clarity about the next question

You do not need a dramatic verdict for the review to help. The evidence may support one focused question, reveal that important context is missing, or show that the concern rests on an ambiguous clue.

Sources and limits

The Cheater AI product team uses these public sources to explain factual limits responsibly. They do not imply clinical, therapeutic, legal, or other professional review of this guide.