We don't tell you what to think.
We show you the evidence.
Chronicle puts every news claim on trial. AI prosecution. AI defense. Multiple model jury. Real source tracing. You deliver the verdict.
Chronicle doesn't check facts the way you're used to.
We don't rely on consensus. We don't trust repetition. A claim repeated a thousand times isn't more true than one said once.
Here's what makes this different
⚔️
Adversarial analysis — Every claim faces both prosecution AND defense. AI models argue both sides.
🗳️
Multi-model deliberation — No single AI decides. Multiple models from different providers deliberate independently, reducing individual bias.
💰
Incentive mapping — We trace who benefits if a claim is true, and who benefits if it's false. Follow the money.
🔍
Source verification — We distinguish between peer-reviewed research, independent reporting, single-source claims, and hearsay. Not all evidence is equal.
⚖️
No editorial bias — No human editorial board decides what's true. The evidence argues for itself.
Open a New Case
Trending Cases
No cases yet. Be the first to open a trial.
Case Archive
How Chronicle Works
Chronicle is not a fact-checker. It's a courtroom. Every claim submitted to Chronicle goes through an 8-agent adversarial analysis pipeline:
- Intake — Extracts and classifies testable claims from the content.
- Source Hunter — Traces the citation chain. Counts truly independent sources (not 50 outlets quoting the same wire story).
- Prosecution — Builds the strongest possible case AGAINST the claim, even if it seems obviously true.
- Defense — Builds the strongest possible case FOR the claim, even if it seems obviously false.
- Historian — Finds historical precedents for similar claims and how they resolved over time.
- Incentive Analyst — Maps who benefits if the claim is true vs false. Follows the money on ALL sides.
- Multi-Model Jury — Runs the claim through Claude, GPT-4o, Gemini, and a skeptic model independently. Where they diverge = signal.
- Synthesis Judge — Combines everything into the Chronicle Report with an uncertainty rating.
The prosecution and defense use different AI models to ensure genuine perspective diversity, not just rephrased agreement.
Chronicle does not render verdicts.
We present the trial. You are the judge.