Skip to content

Image: reversinglabs.com · rights & removal

Executive Summary

An attack campaign, OPTIMJACK, leveraged a vulnerability in Optimizely’s Web Experimentation platform to inject malicious JavaScript into numerous customer websites. This exploitation bypassed traditional security perimeters by distributing the malicious code through seemingly trusted third-party snippets. The payload was capable of various actions, including capturing payment card details from first-party forms, collecting non-HttpOnly cookies, and attempting to deliver exploit kits to mobile users. The infection affected a large scope, spanning over 178 organizations and including major entities like Zoom and Salesforce.
The incident highlights a critical trust gap: even when software comes from a reputable provider and loads via trusted domains, the code it delivers can be maliciously altered. While Optimizely confirmed fixing the vulnerability and removing injected code, public reporting indicates the mechanism enabling cross-project modification remains unclear. Furthermore, the effectiveness of protective measures like Content Security Policy (CSP) varied widely across affected sites, suggesting that defense-in-depth is crucial, but CSP alone is insufficient against malicious changes within approved integrations.

Facts Only

Attackers utilized a vulnerability in Optimizely’s Web Experimentation platform on November 16, 2026.
The attack campaign was named OPTIMJACK by independent researcher Amirul Cyber.
Report URI researcher Scott Helme identified 188 infected snippets across 178 organizations.
Affected organizations included Zoom, CNN, Salesforce, GitLab, IKEA, and HSBC.
The malicious code was delivered via Optimizely snippets, not by compromising application source code or web servers.
The payload could capture payment-card details, replace cryptocurrency wallet addresses, and collect non-HttpOnly cookies.
The loader was inserted into the Project JavaScript configuration of affected projects.
The attack was first observed in a BBC snippet on November 16, at 04:21 UTC.
The payload host 1688360[.]cc was suspended on November 17 at 00:37 UTC.
A later payload version included a card skimmer that targeted fields in first-party forms.

Full Take

<SKEPTICALMODE>
The narrative presented frames a sophisticated supply-chain vulnerability in a trusted software service as a definitive demonstration of a systemic failure of established digital trust models. The strongest version of this narrative is that vendor reputation and initial security assessments are insufficient safeguards; ongoing, deep visibility into dynamically delivered code is a prerequisite for operational security. Report URI meticulously documents the scope and technical capabilities of the attack, detailing payload evolution and the efficacy—or lack thereof—of various browser-side controls like CSP.
A structural concern emerges in how the central problem is framed. While the incident demonstrates a breakdown in the 'trust-but-verify' model inherent in modern web architecture, the piece ultimately pivots toward promoting the necessity of sophisticated investigative tooling (ReversingLabs services). This functions as a subtle lure, where the highly technical description of the exploitation serves to establish a comprehensive "crisis" that only the presented commercial solution can resolve. The assertion that CSP is not a substitute for monitoring, while technically sound, serves to elevate the need for active, artifact-level forensic investigation, a service heavily marketed alongside the technical description.
The underlying assumption driving this entire discourse is that the current reliance on "approved" software is inherently fragile because it is dynamic. The cost borne is primarily the regulatory and reputational risk for affected organizations, while the benefit accrues to security vendors who provide the necessary post-exposure investigation capabilities.
The key question centers not on how the code was delivered, but how the attackers gained the *write access* to so many disparate customer projects without leaving evidence of that initial access vector. What architectural weaknesses allow a single vulnerability in a central provider to translate into unchecked, widespread modification capabilities across an entire client base?
If this narrative were part of an influence campaign, the playbook would involve demonstrating a complex, undeniable security failure (the OPTIMJACK scope) followed by an immediate, solution-oriented prescription (investigative capability). The content aligns closely with this pattern, using the detailed technical exposition not just to inform, but to define the scope of necessary, expensive remediation.
Patterns detected: ARC-0151 Vendor Advertorial, ARC-0044 Appeal to Authority
</SKEPTICALMODE>

From the original · ReversingLabs Blog

Spectra Assure Free Trial Get your 14-day free trial of Spectra Assure for Software Supply Chain Security Get Free TrialMore about Spectra Assure Free TrialOn November 16, 2026, attackers used a vulnerability in Optimizely’s Web Experimentation platform to inject malicious JavaScript into customer projects.
Read the full story at reversinglabs.com

Sentinel unavailable

The automated check of the source article's wording did not complete, so no result is shown.

This looks only at the wording of the original source article, not at this page's AI-written sections. A small local AI model made this estimate. It has not been checked against known human and machine texts, so treat it as provisional. It cannot show who wrote an article.

OPTIMJACK: When Trusted JavaScript Becomes a Supply Chain Attack | Huntaegis