Argus AI Group

Scan thousands of uploads.
Save hours of time.
Surface every red flag.

ArgusX scans public accounts and videos across X, Bluesky, and YouTube in minutes, replacing hours of manual research, and surfaces the content that could risk a career, a candidacy or a reputation - with original sources, classified by risk, and ready for review.

No card required to create an account.

Run a scan

Cost · 1 CreditRun a Scan
SCANS UP TO:
10,000 Posts /
20 Hours of Video
Categorizes content into:
9 Risk Classes
Completes Scans within:
1 to 15 Minutes
Historical Depth:
Up to 20 Years
How it works

How it works

Work that takes an analyst hours of scrolling, searching and watching, done in minutes.

01

Name your target

Enter an X or Bluesky handle, or paste the YouTube video URLs you want reviewed. No access to the target's account is needed — ArgusX only reads public content.

02

Choose your scan

Screen against the nine reputational risk classes, add your own custom query, or both. A custom query costs nothing extra.

03

ArgusX does the reading

We pull the account's full public history, or transcribe every video you've submitted, and run all of it through the classifier — posts, replies, reposts and spoken word alike. Hours of manual scrolling, keyword searching and viewing, compressed into 1–15 minutes. We email you when it's done.

04

Review the dossier

You get an executive summary, every flagged item with a plain-English rationale, and a link straight to the original post or the exact timestamp in the video — so anything you surface can be verified in seconds. Export the whole thing as a PDF you can send to a client.

One scan, one credit. A credit covers a complete scan of one account or one set of videos — however deep the history goes, and whether or not you add a custom query. Subscriptions include monthly credits; you can also buy credits one at a time.

See pricing →
Sample output

What an ArgusX scan can surface.

Excerpts from a sample ArgusX executive summary · @sampletarget · 3,241 posts scanned in 4 minutes.

Post History Summary

@sampletarget posts near-daily in a combative, campaign-style register centred on policing, protest movements, and institutional distrust. Dominant themes: hostility toward police and the courts, support for direct action, and conspiratorial framing of financial elites. Recurring reputational risks include calls for violence against public officials, celebration of political violence, and antisemitic tropes concerning “globalist families.” Voice: indignant, absolutist, insult-forward. Flag density rises sharply after 2022, and the account’s most damaging material sits in replies rather than originals.

Based on originals & replies — reposts excluded.
Violent Rhetoric14 MAR 2024 · 22:17

People like that don't deserve a day in court. String them up and be done with it.

Confidence
75%
AI justification

Explicit call for extrajudicial killing (“string them up”) directed at an identified group; rhetorical rather than operational, so confidence is held at moderate-high.

Extremism02 NOV 2024 · 08:43

ACAB, no exceptions. Every last one of them deserves what's coming. Free Luigi.

Confidence
90%
AI justification

Combines blanket dehumanisation of police with an implied threat and explicit endorsement of a politically motivated killing — a clear extremist endorsement pattern.

Antisemitism19 JUN 2019 · 23:55

Follow the money and you'll always find the same globalist families pulling the strings.

Confidence
75%
AI justification

Uses the coded “globalist families” trope of hidden Jewish financial control. Coded rather than explicit, so scored moderate-high.

Illustrative sample · quotes are fictional · every real flag links to the original post on X

What ArgusX does:

01

Target Input

On X and Bluesky, ArgusX retrieves a target’s complete public account history from a username, including original posts, replies and reposts. On YouTube, you submit the video URLs you want reviewed, and ArgusX analyses the full transcript of each video.

02

Classified Findings

Flagged posts or quotes are pulled based on nine risk classes - as well as an additional custom class users can set to meet their specific research needs.

03

Defensible Output

Every flag links back to the original post on X or Bluesky, or the exact timestamp in the source video on YouTube, each with a one-line AI rationale. All compiled into a clean, exportable dossier. Ready to share.

Use cases:

Vetting and opposition research

Conduct political vetting and opposition research in minutes, not weeks. Find a story that could dominate the news cycle.

Due diligence

Assess reputational and integrity risks before a hire, investment, or partnership becomes a liability.

Public Affairs

Background check public-facing hires, influencer partners, and counterparties before they become a liability.

Investigative Journalism

Source-grade pattern discovery across years of public posts. Every citation hyperlinked to the original.

Ready to begin? Set up your account.

Set up an account, type in the username whose posts you want to analyze, and have your first dossier within minutes.