The TikTok Fact Checker uses AI to analyze public TikTok videos and return web-grounded fact-check triage — verifiable claims, verdicts, cited source URLs, risk levels, and summaries from both spoken audio and on-screen text.
What does TikTok Fact Checker do?
Paste TikTok video URLs and get AI-assisted claim analysis for trust-and-safety, newsroom, and compliance workflows:
- Downloads public TikTok videos (no login)
- Reads post metadata (author, post date, caption, hashtags) to interpret relative wording like "today" or "this morning"
- Watches the full video — speech and on-screen text
- Extracts verifiable factual claims with timestamps
- Verifies each claim with live web search (OpenRouter web search tool)
- Returns verdicts:
true,false,misleading,unverifiable, oropinion - Attaches cited source URLs with titles and snippets per claim
- Scores overall misinformation risk (
low/medium/high)
Disclaimer: This is AI-assisted triage with web search, not professional fact-checking. Verdicts are model judgments grounded in retrieved web sources. Verify independently before publishing or enforcement decisions.
What data can you extract?
| Field | Description |
|---|---|
videoUrl | Canonical public video URL |
videoId | TikTok item id |
duration | Video length in seconds |
videoMetadata | TikTok post context: author, postedAt, caption, hashtags, engagement, music |
claimCount | Number of claims identified |
overallRisk | Overall misinformation risk (low / medium / high) |
summary | 2–3 sentence overview of the video credibility |
claims | Array of claim objects (see below) |
disclaimer | AI limitation notice |
checkedAt | Analysis timestamp (ISO) |
Each claim in claims:
| Field | Description |
|---|---|
claim | The factual statement in plain language |
timestampSec | Approximate second when the claim appears |
source | spoken, on_screen, caption, or mixed |
videoEvidence | What was spoken or shown on screen in the video |
verdict | true, false, misleading, unverifiable, or opinion |
confidence | high, medium, or low |
rationale | Brief explanation grounded in web sources |
riskLevel | Harm risk if the claim is wrong (low / medium / high) |
sources | Array of { url, title, snippet } web citations |
searchQueries | Search queries used to verify this claim |
videoMetadata object:
| Field | Description |
|---|---|
author | TikTok username |
authorId | TikTok user id |
postedAt | Video post date (ISO) — used to interpret "today" / "this morning" in claims |
description | Caption / on-post description |
hashtags | Hashtags from caption or tags |
viewCount | View count when downloaded |
likeCount | Like count when downloaded |
commentCount | Comment count when downloaded |
shareCount | Share/repost count when downloaded |
musicTitle | Audio track title |
webpageUrl | Canonical TikTok page URL |
Sample output
{
"videoUrl": "https://www.tiktok.com/@aaronrheins/video/7688742727378291999",
"videoId": "7688742727378291999",
"duration": 25,
"videoMetadata": {
"author": "aaronrheins",
"postedAt": "2024-03-26T12:00:00.000Z",
"description": "RAF Hawk T2 crash in Wales — both pilots ejected",
"hashtags": ["news", "raf", "breaking"],
"viewCount": 1200000,
"likeCount": 45000,
"commentCount": 890,
"shareCount": 1200,
"musicTitle": "original sound - aaronrheins",
"webpageUrl": "https://www.tiktok.com/@aaronrheins/video/7688742727378291999"
},
"claimCount": 12,
"overallRisk": "medium",
"summary": "This video reports on a Royal Air Force jet crash shortly after takeoff from RAF Valley. Both crew members ejected; eyewitnesses reported flames on takeoff.",
"claims": [
{
"claim": "A Royal Air Force jet crashed this morning.",
"timestampSec": 1.5,
"source": "spoken",
"videoEvidence": "Royal Air Force training jet crashed this morning",
"verdict": "true",
"confidence": "high",
"rationale": "BBC and ITV reported an RAF Hawk T2 crash in Wales on the video's post date; 'this morning' refers to that day, not the fact-check run date.",
"riskLevel": "low",
"sources": [
{
"url": "https://www.bbc.com/news/uk-wales-68664078",
"title": null,
"snippet": null
}
],
"searchQueries": ["RAF Hawk T2 crash Wales March 2024"]
},
{
"claim": "The plane involved was a Hawk T2 departing RAF Valley.",
"timestampSec": 22.2,
"source": "spoken",
"videoEvidence": "The plane was a Hawk T2 departing RAF Valley",
"verdict": "true",
"confidence": "high",
"rationale": "Multiple outlets identify the aircraft as an RAF Hawk T2 that departed RAF Valley.",
"riskLevel": "low",
"sources": [
{
"url": "https://www.itv.com/news/wales/2024-03-26/royal-air-force-jet-crashes-in-wales-with-two-pilots-ejecting",
"title": null,
"snippet": null
}
],
"searchQueries": ["RAF Hawk T2 RAF Valley crash"]
}
],
"checkedAt": "2026-09-24T08:45:00.893Z"
}Use the Claims dataset view in Apify Console for one row per claim (verdict, rationale, sources).
Why fact-check TikTok videos?
- Newsroom triage — prioritize videos with high-risk health or political claims
- Trust & safety — flag misleading product or financial promises before they spread
- Brand safety — screen creator content for false claims before sponsorship
- Compliance monitoring — audit influencer claims against advertising standards
- OSINT workflows — cross-check breaking-news clips with cited web sources
- Election integrity — surface misleading political statements in short-form video
- Health misinformation — flag unverified medical or supplement claims early
- Creator verification — compare spoken claims against the post date and caption context
How to use TikTok Fact Checker
- Open the Actor in Apify Console
- Paste one or more public TikTok video URLs
- Set Max videos to check (default 5)
- Keep US Residential proxy enabled (required — TikTok blocks datacenter IPs)
- Click Start
- Download results from the Output tab (JSON, CSV, Excel, or API)
Example input:
{
"videoUrls": [
"https://www.tiktok.com/@aaronrheins/video/7688742727378291999"
],
"maxItems": 3,
"proxyConfiguration": {
"useApifyProxy": true,
"apifyProxyGroups": ["RESIDENTIAL"],
"apifyProxyCountry": "US"
}
}Input
| Field | Required | Default | Description |
|---|---|---|---|
videoUrls | yes | — | Public TikTok video URLs |
maxItems | no | 5 | Max videos per run |
proxyConfiguration | no | US Residential | Proxy for video downloads (required) |
How much does it cost?
This Actor uses pay-per-event + platform usage billing:
| Event | What you pay for |
|---|---|
video-checked | Each TikTok video successfully fact-checked |
video-second | Each second of video processed (rounded up) |
| Platform usage | Apify compute and residential proxy (passed through) |
Short news clips (~30s) typically cost a few cents in platform usage plus the event fees. Start with maxItems: 1 to estimate cost for your workflow.
Integrations
- API — call from any language via the Apify API
- Zapier / Make — trigger fact-check runs from spreadsheets or alerts
- Webhooks — notify Slack or email when a high-risk video is flagged
Image Credit
Image credit: TikTok

