Frequently Asked Questions

ArgusX is an automated reputational risk screening tool.

You give it a public X (Twitter) account, and it pulls that account's post history — original posts, retweets, and replies — then runs every item through an AI classification pipeline that flags content carrying potential reputational risk. Instead of manually scrolling through thousands of posts, you get a structured report highlighting the posts most likely to matter, so you can focus your attention where it counts.

By default, ArgusX screens against nine reputational risk categories: racism, violent rhetoric, antisemitism, Islamophobia, transphobia, homophobia, sexism, hate speech, and extremism.

Every post, retweet, and reply is assessed against each of these, and anything that trips a category is surfaced in your report with the relevant flag attached.

Beyond these nine, ArgusX offers a tenth, user-defined class — the Custom Query. Sometimes the standard reputational categories aren't what you're worried about. You might not care about general red flags at all; instead you need to know what an account has said about a specific issue, person, organisation, or set of topics.

A Custom Query lets you define exactly what you're looking for, and ArgusX classifies the account's history against that brief. It's useful for screening an individual's stated views on a particular policy, checking for past commentary on a company or counterparty, or surfacing everything an account has said across a defined range of subjects.

You can add a Custom Query to a standard scan free of charge — it runs alongside the nine reputational categories at no extra cost. Alternatively, you can switch the nine standard categories off entirely and run a Custom Query on its own. Either way, the scan costs a single credit.

Together, that makes ArgusX suited to due diligence, vetting, compliance, HR, and communications work where you need to understand someone's public posting history quickly and defensibly.

One scan costs one credit. Credits are sold in three packs, and you can pay in GBP, USD, or EUR:

The larger the pack, the lower the per-scan cost — from £69 / $99 / €84 on the Starter pack down to roughly £62 / $85 / €72 per scan on Enterprise. One credit covers one scan regardless of whether you run a FastScan or a DeepScan.

Note: Prices may change over time, so please consult the billing section for the most up to date prices.

Once you've created an account, go to the Billing section of your dashboard, choose your pack and currency, and check out. Your credit balance updates immediately and is shown at the top of the screen.

Credits are deducted one per scan as you use them.

Both run the same risk classification — the difference is how much history they pull and how long they take.

FastScan pulls an account's most recent ~3,200 posts, retweets, and replies, and completes in roughly 0–5 minutes. It's the right choice when you need a result urgently, or when the account has fewer than 3,200 posts — in which case FastScan captures effectively everything.

DeepScan takes longer but can reach much further back. It typically returns somewhere between 5,000 and 14,000 posts, though the exact number depends on factors outside our control — chiefly how much traffic X is handling at the time. At a minimum it returns as much as a FastScan would; usually it returns considerably more. DeepScans normally take 10–15 minutes, but on very large accounts can run to half an hour or more.

In short: FastScan for speed and smaller accounts, DeepScan for depth on accounts with long histories.

Check the account's post count first. On an X profile, the total number of posts is shown directly beneath the account name at the top of the profile.

If that number is under 3,200, a FastScan will capture effectively the entire account and is the better choice. If it's well above 3,200, a DeepScan is worth the extra time to reach further back into the history.

One caveat: the post count shown on a profile can include private or otherwise restricted content. ArgusX can only access public posts, so the number we return may be lower than the total displayed on the profile. This isn't an error — it reflects that some content simply isn't publicly available to scan.

This is by design. We've deliberately tuned the classifier to err on the side of sensitivity, because the cost of missing a genuinely damaging post is far higher than the cost of surfacing a borderline one. As a result, you'll occasionally see posts flagged — for example under a category like "racism" — that, on reading, you don't consider a real concern.

That's expected, and to a large extent unavoidable: what counts as a "red flag" is partly subjective and context-dependent, and no automated system can make that judgment perfectly. What ArgusX does is strip out the overwhelming majority of posts that are highly unlikely to be problematic, leaving you a far smaller set to review with human judgment. Even with a few false positives, that saves hours of manual reading per account. Think of the flags as a prioritised shortlist for your attention, not a final verdict.

We want to hear it. You can email us at info@argusgroup.ai, or use the form in the Help section of your dashboard — just enter your email and your inquiry or feedback and hit send. Feature requests, bug reports, and general suggestions are all welcome and genuinely shape what we build next.