A hand holding a phone running the What the Fact app over a social media feed

Fact-check your feed instantly.

Share a post, get the truth back in seconds.

  • Cross-checked against 6+ independent, vetted outlets
  • AI-generated media detection built in from day one
  • Free to download, with a meaningful free tier at launch
The problem

The internet already fell for these.

Odds are, so did you. Misinformation doesn't look like misinformation. It looks like a real photo, a trusted-sounding account, or a claim that's technically true.

The shark that “swam onto the highway”

A shark appearing to swim down a flooded highway next to a car's wing mirrorRECYCLED

The same flood photo gets recycled every hurricane season with a new city's name on it. It's real, just not from this storm, and not from this year.

AnalysisReal photo, fake news

The chocolate experience that never existed

Willy's Chocolate Experience promo — a vivid, AI-generated candyland with a 'Book Now' buttonAI-GENERATED

AI-generated marketing images sold out a “Willy Wonka” event in Glasgow in 2024. Families showed up to an empty warehouse. The photos looked real enough to book tickets.

AnalysisSuspicious account

Hotdogs and cigarettes have the same cancer risk

Viral post: pans of bacon, sausages and ham over the caption 'classified in the same cancer risk as cigarettes'

Processed meat and cigarettes are both classified as level-one carcinogens, but cigarettes multiply cancer risk by as much as 20×, far worse than an occasional hotdog.

AnalysisTechnically true, but misleading

The claim, the account, the media: a post can go wrong on any of them, and checking takes longer than sharing. What the Fact checks all three in seconds.

How it works

The claim, the account, the media. All checked in seconds.

  1. 01

    See something suspicious? Share it.

    Direct from your feed on X, Facebook, Instagram, Threads, LinkedIn, Reddit, YouTube, or TikTok.

  2. 02

    What the Fact begins the verification.

    The claim, the account, and the media all get checked at once: against 6+ independent sources, for AI generation and manipulation, and automation signals. No algorithmic opinions, just receipts.

  3. 03

    Get your analysis.

    A clear analysis, a confidence score, and every source behind it, ready to share and debunk misinformation.

Daily fact checks

Trending claims, checked against the record.

Twice a day we publish fact-checks on the claims making the rounds — the verdict, the reasoning, and the receipts.

Browse the archive→

Did 60 Minutes find Iran hit 31 U.S. military and diplomatic sites in eight countries?

X · 60 Minutes post 7.1K likes; Raw Story, Daily Beast, NY Post (whistleblower angle) 16 Nov; Pentagon called a related report 'disgusting'; Trump attacked the report on Truth Social 16 Nov 14:17Z ('99.99%' of Iranian missiles and drones stopped)

True

This is the genuine CBS 60 Minutes segment broadcast on 15 November 2026, and its central claims — 31 U.S. sites hit across eight countries, destruction of the Fifth Fleet headquarters in Bahrain, 23 troops killed, and continued attacks after Trump's 1 April victory declaration — are corroborated by Pentagon data and wire reporting. The caption accurately represents what the broadcast reported.

ConflictPoliticsView details→

Did the US attack Tehran with a powerful missile, as this video shows?

X · @Americatwisst 1.4K likes; identical copy-paste posts by several accounts through 15 Nov; Misbar fact-check 16 Nov

False

The video is genuine footage, but it does not show a US attack on Tehran: it records the 9 June 2026 gas distribution station fire at Novy Sulak in Dagestan, Russia, and no US missile strike on Tehran occurred around 16 November 2026.

ConflictView details→

Were dozens of fighter jets deployed to protect Trump from Iran?

X · @Iran_War24 598 likes; re-posted 16 Nov by other accounts; Misbar fact-check 16 Nov

Misleading

The clip is genuine but a year old: it shows six Egyptian F-16s ceremonially escorting Air Force One into Sharm el-Sheikh on 13 October 2025, not 'dozens' of jets deployed to shield Trump from Iran in November 2026. The caption recycles real footage with a false date, reason and aircraft count.

ConflictPoliticsView details→
FAQ

Questions, answered.

Everything you'd want to know before you download.

What is What the Fact?+

What the Fact is a free fact-checking app for iOS and Android. Share a social media post or video to it and it tells you whether the claim is accurate or misleading, whether the account is trustworthy, and whether the image or video is real or AI-generated, with a confidence score and links to every source used.

How much does it cost?+

What the Fact is free to download and start checking. Any premium options will always leave a meaningful free tier. The goal is to make fact-checking accessible, not gated.

Which platforms can I check?+

Share or paste a link from X, Facebook, Instagram, Threads, LinkedIn, Reddit, YouTube, or TikTok, straight from the share sheet. You can also check a post you've screenshotted when a link isn't available.

Is my data private?+

We analyze the public post you share, not your accounts or your contacts. Your verification history is yours, saved so you can revisit past checks.

What sources do you check against?+

Every claim is weighed against 6+ independent, pre-vetted outlets and existing fact-checks. An outlet only counts toward the analysis after we've evaluated its history: how long it's been publishing, how often it posts, and what it actually publishes. Every analysis links to the sources it used.

How does the analysis actually work?+

Three checks run on every post: the claim is cross-checked against vetted coverage; the posting account is profiled for its age, posting pattern, and automation signals; and the media is fingerprinted and run through synthetic-media detection models that flag AI generation and manipulation. The three signals combine into one clear answer with a confidence score.

How accurate is the analysis?+

Every analysis comes with a confidence score and a full signal breakdown, and links to the sources behind it. When the evidence is genuinely mixed, we mark a post Can't Verify rather than guessing.

What does 'Can't Verify' mean?+

It means the claim can't be confirmed or debunked with confidence right now, often because sources disagree or the story is still developing. It's a signal to wait, not to trust or dismiss.

Can it detect deepfakes and AI-generated media?+

Yes. Media is run through synthetic-media detection models that flag AI-generated images and video, edits, and manipulations. It's a core part of every check.

A fact-checker in your pocket.

Know what's real before you share it. Free to download on iOS and Android.

Free to download · no payment required