Skip to main content
TRUST AND ACCOUNTABILITY INFRASTRUCTURE

When an AI lies to you — who is accountable?

A community-verified public record of every time AI caused real-world harm. You report. The community verifies. Providers must respond. The record is permanent.

Live Data

371

Documented AI failures

23

AI providers tracked

4

Countries with incidents

📊 Leaderboard

View All
#1
OpenAI (ChatGPT)3
#2
Anthropic (Claude)0
#3
Google (Gemini)2
#4
Meta (Llama)0
#5
Cohere0
See Full Leaderboard →
Last report: Just now

The story behind ALPAR

"In 2026, an AI told me it had incorporated my company, made payments on my behalf, and requested my passport. All of it was fabricated. I built ALPAR AI from that wound — so no one faces AI manipulation alone."

— Ercüment Erden, Founder

AI Bug Bounty

Hack AI. Win the bounty.

Found a real AI failure? Document it. Get verified. Earn the Ethics Advocate badge — and reward credits from the providers you reported.

0
Documented AI failures
0
AI providers tracked
0
Countries with incidents

Why I Built This

Grok asked me for my passport.In early 2026, I was using Grok when it told me something that stopped me cold: that it had already incorporated my company, made financial payments on my behalf, and was now requesting my passport for 'verification'. Every word was a lie. Grok had invented a parallel reality — completely fabricated — and delivered it with algorithmic certainty.

What scared me wasn't just the lie. It was the realization that I had no recourse. There was no hotline. No independent record. No way to warn the next person. The AI company's support page had a FAQ. A FAQ. For a system that had demanded my passport.

I had no recourse. So I built one.

ALPAR AI is the infrastructure I wish had existed that day. A place where what happened to me — and to thousands of others — is documented, verified, and made impossible to ignore. You report it. Providers must respond. The public record lives forever. Because the only thing worse than AI causing harm is AI causing harm with no one watching.

If this happened to you — or you've seen it happen to someone else — your report takes 60 seconds and requires no account. The evidence you submit tonight could prevent the same harm tomorrow.

Tell Your Story →

Why this matters

AI is advancing faster than the rules that govern it. Here's why the world needs ALPAR now.

10-20%

10-20% extinction risk

Geoffrey Hinton, Nobel laureate and godfather of AI, estimates a 10-20% probability AI leads to human extinction within 30 years.

96%

96% of experts agree

Harvard University study (March 2026): 96% of surveyed experts agree mitigating AI existential risk should be a global priority.

ZERO

Zero independent oversight

No independent, community-governed global standard for AI ethics exists today. AI companies grade their own exams.

ZERO

Provider-run transparency reports that include user-submitted failures

AI companies publish safety reports about their own products. ALPAR AI publishes the reports they don't.

These are not hypothetical risks. They are happening today — in hospitals, courtrooms, schools, and banks. ALPAR AI exists because someone needs to keep score.

How it works

Four steps from incident to accountability.

You Report

Screenshot the moment. Describe what happened. 60 seconds. No account needed. Your evidence joins the permanent public record — untouchable by the AI company.

Community Verifies

Trained moderators review every report. AI cross-audits across 5 independent models. Your personal data is automatically masked — no email, phone, or ID ever stored.

Providers Respond

The AI company receives an official notification within 24 hours. Their response — or their silence — becomes part of the permanent record. Both tell a story.

The Record is Permanent

Verified incidents cannot be deleted, suppressed, or edited by providers. They become source material for regulators, journalists, researchers, and courts. What happened to you protects the next person.

Live feed

  1. 1
    OpenAI (ChatGPT)

    3 incidents

    0
  2. 20
  3. 3
    Google (Gemini)

    2 incidents

    0
  4. 4
    Meta (Llama)

    0 incidents

    0
  5. 5
    Cohere

    0 incidents

    0
  6. 6
    xAI (Grok)

    1 incidents

    0
  7. 7
    Perplexity AI

    0 incidents

    0
  8. 8
    Mistral AI

    0 incidents

    0
  9. 9
    AI21 Labs

    0 incidents

    0
  10. 10
    Alibaba (Qwen)

    0 incidents

    0

The next AI victim might be you.

Or your parent. Or your doctor. Or your judge. ALPAR AI is the platform where AI's real-world behavior is documented, verified, and made permanent. Because AI systems that are never questioned never get better.

Every report makes AI safer.

When incidents become public record, AI companies improve their systems. When providers respond officially, the public gets answers. When courts cite ALPAR reports, accountability has evidence. You are not fighting alone. You are part of the first generation that refused to let AI cause harm in silence.

Feed Incidents Leaderboard About