August 30, 2022 Source: Threatpost 3 min read · 622 words

Watering Hole Attacks Push ScanBox Keylogger

Атаки на водопої поширюють keylogger ScanBox

When It Started—And How Long It's Been Going

August 2022. That's when researchers first reported an active campaign that's been quietly running in the background, harvesting data from unsuspecting website visitors. The timeline matters here because it suggests the attack had already been running before anyone publicly caught it. Security researchers working with Threatpost uncovered what amounts to a surgical operation: a watering hole attack attributed to APT TA423, a sophisticated threat actor known for precisely targeted operations.

Watering hole attacks aren't new. But this one is different. It's not theoretical. It's not a lab exercise. Real people visiting compromised websites have been exposed to the ScanBox keylogger—a JavaScript-based reconnaissance tool designed to steal credentials and monitor activity.

The Discovery

Threatpost reported the findings, and here's what caught researchers' attention: the malware wasn't hiding on some dark corner of the internet. It was injected into legitimate websites—places people visit every day, believing they're safe. That's the whole point of a watering hole attack. You don't attack the defense. You attack the watering hole where your targets come to drink.

The malware itself is built in JavaScript, which means it executes in the browser without requiring installation. No popup asking for admin privileges. No executable file sitting on disk. Just code running invisibly in the background.

Researchers identified the infrastructure, traced the command-and-control connections, and attributed the campaign to APT TA423 based on operational patterns and targeting similarities with known APT cyber attack examples from this group.

Technical Analysis

So what exactly is ScanBox doing?

It's a reconnaissance and data-stealing tool. The keylogger component captures everything typed—passwords, search queries, credit card numbers, whatever. But it does more than that. It also collects browser information, system details, and network data. The tool's designed to be lightweight and persistent, sitting quietly on infected systems while exfiltrating data to remote servers controlled by the attacker.

Here's the part that stings: JavaScript-based malware like this slips past many traditional security tools. It doesn't trigger antivirus alerts the way an executable would. It's not a file you can quarantine. It's code executed in memory, living and dying within the browser session—unless the infection persists across sessions through malicious scripts.

The attack chain works like this. A visitor lands on a compromised website. Malicious JavaScript loads silently. The ScanBox keylogger initializes. And suddenly, every keystroke belongs to the attacker. The real question is: how many websites were actually compromised, and how many visitors does that represent?

Damage Assessment

That's the question Threatpost and other researchers are still trying to answer with precision. What we know is this: APT TA423 isn't running spray-and-pray campaigns. They're selective. Their targets matter. That suggests the damage might be concentrated among specific organizations or sectors rather than broadly distributed.

But concentrated doesn't mean small. APT cyber crime operations targeting specific victims often inflict severe damage—data breaches, espionage, financial theft, intellectual property loss.

Threatpost didn't disclose which websites were compromised, likely for good reasons: some might still be infected, and public disclosure could warn the attackers to move faster. But the absence of detail also means security teams can't easily determine if their web properties were among the targets.

Mitigation

If you manage a website, here's what matters:

Audit your web infrastructure. Check for unauthorized code injections. Review your content delivery systems and any third-party scripts you're loading. This is particularly nasty because attackers can hide malicious code in legitimate-looking libraries or plugins.

For end users: keep browsers updated. Use content security policies. Consider browser extensions that block scripts from unknown sources. And frankly, assume any website could be compromised—use unique, strong passwords everywhere.

Organizations should monitor for unusual JavaScript execution patterns and consider restricting third-party script loading where possible. The APT vulnerability here isn't a software bug you can patch. It's an operational security gap. That requires constant vigilance and architecture that assumes compromise.

Read original article →

// FAQ

What is ScanBox and what does the keylogger do?

ScanBox is a JavaScript-based reconnaissance and keylogging tool that captures everything typed in browsers, collects system information, and exfiltrates data to attacker-controlled servers. It executes in memory without requiring installation.

How do I know if I visited a compromised watering hole website?

There's no easy way to verify retroactively. Monitor your accounts for suspicious activity and consider changing passwords for critical services. If you visited unknown websites in late 2022, assume possible compromise.

Is APT TA423 the same as other known APT groups?

No. APT TA423 is a distinct threat actor group known for targeted campaigns. Researchers attributed this specific watering hole campaign to them based on operational patterns and infrastructure analysis rather than shared tactics with other APT groups.

Concerned about your project's security? Run an automated pentest with AISEC — AI-powered scanner with expert verification. Go to dashboard →