Medium Severity

Evasion: Hidden text using CSS-obscured HTML option labels

Labels

Credential Phishing
Evasion
Social engineering
Impersonation: Brand
HTML analysis
URL analysis
Content analysis

Description

Detects inbound messages containing HTML elements that are visually hidden through CSS techniques such as extreme negative z-index values or zero-sized dimensions, combined with <option> tags containing label attributes. This pattern is indicative of hidden text salting, where invisible or off-screen content is injected into the HTML body to manipulate content analysis engines and evade detection.

References

No references.

Sublime Security
Created Sep 1st, 2026 • Last updated Sep 1st, 2026
Feed Source
Sublime Core Feed
Source
GitHub
type.inbound
and any(html.xpath(body.html, "//option[.//table or .//article]").nodes,
        any(ml.nlu_classifier(.inner_text).intents,
            .name == 'cred_theft' and .confidence != 'low'
        )
)
MQL Rule Console
•Docs•Learning Labs

Playground

Test against your own EMLs or sample data.

Share

Post about this on your socials.

Get Started. Today.

Managed or self-managed. No MX changes.

Get Started
Deploy and integrate a free Sublime instance in minutes.
Get Started