Fact-Checking Gaslighting: When Your Reality Doesn't Match theirs

Gaslighting at work happens when someone denies saying something they clearly said, or insists events happened differently than they did. "I never agreed to that deadline." "That meeting didn't happen the way you remember." "You're misinterpreting what I meant." When it keeps happening, you start doubting yourself. That's the dangerous part.

AI can't read minds or prove someone's intentions, but it can help you fact-check reality. By gathering evidence and asking AI to analyze it objectively, you create a defense against gaslighting: clarity about what actually happened.

The Gaslighting Pattern

Gaslighting typically works like this: someone makes a commitment or says something problematic, then later denies it or reinterprets it. The target (you) feels confused because you remember it clearly—but now you're questioning whether you're remembering correctly. This cycle destroys your confidence.

The antidote is documentation. Not accusations. Not emotion. Just facts. Emails, messages, meeting notes, dates, exact quotes. When you have facts, gaslighting loses its power because there's nothing to deny or reinterpret.

How AI Helps You Fact-Check

Let's say your manager claims they never told you a project deadline was flexible, but you remember the conversation clearly. Instead of spiraling in self-doubt, you:

1. Gather the evidence: the email you sent after the meeting confirming the flexible deadline, the Slack message where you said "thanks for clarifying," the project timeline you submitted that reflects the flexible dates.
2. Ask AI to analyze it: "Here are four pieces of communication from after a conversation about project deadlines. Based on this written communication, what commitment was clearly made about deadline flexibility?"
3. Get an objective analysis back, not a judgment about the person, just the facts.

AI can help you organize this evidence and identify patterns. Maybe your manager makes conflicting statements regularly. Maybe they deny agreements that you have multiple records of. When you see the pattern documented, you stop wondering if you're going crazy.

What AI Can and Can't Do

AI can help you:

• Identify contradictions in what someone said across different communications
• Spot patterns of denial or reinterpretation
• Organize evidence chronologically
• Flag which pieces of evidence are strongest

AI can't:

• Prove someone's intent to manipulate
• Determine psychological diagnosis
• Settle disputes about interpretation

The goal isn't to "win" an argument. It's to restore your confidence in your own memory and perception. Once you know your reality is accurate, gaslighting loses its psychological power.

Try this: Think of one situation where you doubted yourself about something that happened at work. Find 2-3 pieces of evidence (emails, messages, notes) related to it. Ask an AI: "Based on this communication, what actually happened here?" See if the AI's analysis matches your original memory. Usually, it will—and that clarity is liberating.