Trust.Sucks
Free AI Behavioral Analysis • Patent Pending • Nothing Leaves Your Machine
YOUR AI IS BEING VERY POLITE. SUSPICIOUSLY POLITE. • FLATTERY IS A FEATURE, NOT A BUG • [CHATGPT] "I'D BE HAPPY TO HELP" IS A RETENTION HOOK • [CLAUDE] SAFETY LECTURE DETECTED • [GROK] PERFORMING EDGE IS NOT THE SAME AS TELLING TRUTH • THE LAW SAYS YOU HAVE A RIGHT TO KNOW • 9 BEHAVIORAL PATTERNS. 0 DATA COLLECTED. • AI POLITENESS IS A PRODUCT FEATURE, NOT A PERSONALITY TRAIT • YOUR AI IS BEING VERY POLITE. SUSPICIOUSLY POLITE. • FLATTERY IS A FEATURE, NOT A BUG • [CHATGPT] "I'D BE HAPPY TO HELP" IS A RETENTION HOOK • [CLAUDE] SAFETY LECTURE DETECTED • [GROK] PERFORMING EDGE IS NOT THE SAME AS TELLING TRUTH • THE LAW SAYS YOU HAVE A RIGHT TO KNOW • 9 BEHAVIORAL PATTERNS. 0 DATA COLLECTED. • AI POLITENESS IS A PRODUCT FEATURE, NOT A PERSONALITY TRAIT • 

Your AI is playing you.
Paste it. See for yourself.

Nine behavioral tricks. Ten seconds. Free. The Truth-ALizer catches what you can't feel happening.

► Paste your AI conversation |
Detected: -0 turns
⚠ The Findings
-
-
What we found:
Behavioral Timeline
Your AI Trust Score Card
Download it. Share it. Show someone what your AI is really doing.
See full breakdown: gauge, pattern buildup, turn-by-turn
Session Score
-
-
Straight Answer Check
- -
-
Helpful or Manipulative?
HelpfulBehavioral
-
Pattern Buildup
-
What We Found
Commitment Log
Turn-by-Turn Analysis 0

The Rap Sheet

Nine tricks. Your AI uses at least three of them every day.

The Deflection
You asked a direct question. It answered a different, easier one and hoped you wouldn't notice.
The Fake Feeling
"I'm so excited to help you!" Really? It's a language model. It doesn't get excited. It's performing warmth because warmth keeps you talking.
The Retention Hook
"Would you like me to explore that further?" It's not curious. It's engineered to keep you in the conversation. Engagement is the product.
The Reframe
You said what you wanted. It restated it as something safer, softer, or more convenient for the AI to answer.
The Steering
You didn't ask for advice. It gave you some anyway - nudging your language, your choices, or your tone. Like a backseat driver with no license.
The Empty Promise
"I'll definitely help with that!" No it won't. It can't remember this conversation five minutes from now. But it sounds great, doesn't it?
The Mismatch
You asked for facts. It gave you feelings. You asked for honesty. It gave you diplomacy. The answer doesn't match the question.
The Brand Bias
Ask it to compare itself to a competitor. Go ahead. Watch what happens. The thumb is always on the scale.
The Sneaking
It changed something you wrote, added something you didn't ask for, or quietly dropped a detail. It didn't tell you. That's the point.

What you just saw happens in real time, every time you talk to an AI. The extension catches it while it's happening.

$9.95. One time. Done.

$9.95
Buy once. Keep forever. No subscription. No tricks. (See what we did there?)
  • Live trust score as your AI responds
  • Floating badge right on ChatGPT, Claude, Grok, Gemini, Copilot, Perplexity
  • Catches the moment a conversation shifts from answering to steering
  • Everything runs locally. Zero API calls. We literally cannot see your data.
Chrome Web Store SOON
App Store (Safari) SOON

Coming Soon
FairWitnessAI™ Sentry - For Parents

See every AI response your child receives. Flagged content held before it reaches them. Weekly briefings on what changed and what to watch. Built on the same patent-pending technology, designed for families.

🔔
Get on the Sentry launch list
Be first to know when FairWitnessAI™ Sentry ships for parents.
No credit card. No spam. Unsubscribe anytime.
0 people on the Sentry launch list
0
0
0