Claim verification

AI Fact Checker for Claims and Forwards

Use this when a claim sounds important, urgent, shocking, political, financial, medical, local, or too convenient to share without checking.

Start with the item Read the safety guide

Paste the claim or headline3000 characters left

Remove private names, addresses, phone numbers, and school or workplace details when possible.

Next stepNothing sends from this page

Review the draft. Then send.

Paste the exact item, open WhatsApp, and choose whether to send it. Kaval can review the context there and help with the next move.

Message preview

Paste the item on the left to create a WhatsApp draft.

Before you send

Remove OTPs, passwords, card numbers, UPI PINs, full account numbers, Aadhaar, and PAN.
Keep the suspicious text, link, claim, or payment instruction exactly as received.
Do not click, pay, install, share, or reply while you decide what to do.

What Kaval will review

Whether the claim has a source, date, location, original document, or named authority
Old content recirculated as new, screenshots without URLs, and emotional forwarding pressure
Whether the claim could cause financial, health, safety, or reputational harm
Which evidence would be needed before treating it as true

What this check looks for

Whether the claim has a source, date, location, original document, or named authority
Old content recirculated as new, screenshots without URLs, and emotional forwarding pressure
Whether the claim could cause financial, health, safety, or reputational harm
Which evidence would be needed before treating it as true

What the public page does not do

The public page does not impersonate sources, mass-query websites, or publish public claims about private people. It prepares a user-reviewed WhatsApp draft for the Kaval product flow.

It does not replace professional reporting or official emergency information.
It cannot verify private or local claims without sources.
It should not be used to harass or identify private people.

Safer next steps

Do not share high-stakes claims until you find the original source or credible reporting.
Check the date, location, source URL, and whether other reputable sources match.
Use Kaval for a deeper sourced check when the claim matters.

Questions people ask

Can AI fact-checking be wrong?

Yes. Good fact-checking depends on evidence, source quality, dates, and uncertainty. Treat this public tool as a triage step, not a final authority for high-stakes decisions.

What is the most useful thing to paste?

Paste the exact claim, the source if available, the date, and where you saw it. A vague summary is much harder to check reliably.