Optimizely Supply Chain Attack Plants Malware Across Major Global Websites
A vulnerability in Optimizely's web experimentation platform allowed attackers to inject malicious JavaScript into the websites of at least 178 major organizations. The compromised scripts, which bypassed standard security defenses by loading from a trusted domain, attempted to steal cryptocurrency, skim payment cards, and deliver iOS exploits.
By Lila Morgan
Share this story
On November 16, 2026, a sophisticated supply chain attack turned one of the internet's most trusted testing tools into a malware delivery network. Attackers exploited a vulnerability in Optimizely Web Experimentation, a platform used by major corporations to test page designs and layouts.[1][2]
The breach allowed unauthorized actors to inject malicious JavaScript into the websites of at least 178 organizations. The compromised list reads like a directory of the global economy, including the BBC, CNN, Zoom, Salesforce, HSBC, and Equifax.[1][4]
Visitors to these sites unknowingly executed a multi-purpose malware kit directly in their browsers. Because the malicious code was served through Optimizely's official content delivery network, it bypassed many standard security filters and operated with the full trust of the host website.[1][4]
Security researchers named the campaign OPTIMJACK, highlighting its broad reach and stealthy execution. The malware was designed to swap cryptocurrency wallet addresses, skim payment card details, and deliver exploit kits to Apple mobile devices.[2][3]
"Earlier today, Optimizely identified a security incident in which an unauthorized party exploited a vulnerability in Experimentation and used it to add malicious JavaScript to some customer projects," the company stated in its incident notice.[5]
How the Injection Worked
The vendor confirmed that it secured the affected user accounts and implemented additional protections to prevent unauthorized code delivery. However, the company has not yet disclosed exactly how the attackers gained write access across so many unrelated customer projects.[2][5]
Optimizely Web Experimentation allows site owners to run A/B tests by loading a small JavaScript snippet from the company's servers. This snippet includes a "Project JavaScript" field, which is designed to let customers run custom code across their entire site.[1][3]
The attackers managed to write a malicious loader into this specific field while leaving the rest of the snippet untouched. Because the tampered files were still delivered from Optimizely's domain and carried the vendor's legitimate cryptographic certificate, browsers treated the code as trusted first-party content.[1][4]
This meant the malware inherited the permissions of the host website, granting it full access to read and modify the page. No individual customer website or internal repository had to be breached for the attack to succeed.[1][2]
The first injection was detected at 04:21 UTC on November 16, appearing in the BBC's Optimizely snippet. Throughout the day, the payload evolved, with the attackers deploying different loader variants, including base64-encoded versions and inline scripts.[1][4]
Cryptocurrency and Card Skimming
The incident highlights a critical blind spot in modern web development, where sites routinely execute third-party code without secondary validation. When a vendor's infrastructure is compromised, the malicious payload is instantly pushed to millions of end users.[2][3]
The OPTIMJACK payload was highly modular, with its most immediate financial impact targeting cryptocurrency users. The script actively monitored the victim's operating system clipboard for cryptocurrency wallet addresses.[1][3]
When a user copied a legitimate address from an affected website, the malware silently swapped it for an attacker-controlled destination before it could be pasted. The script even checked that the copied text was a valid wallet format and looked for adjacent keywords like "deposit" before executing the swap.[1][4]
Initial reports suggested the attackers netted roughly $7,040, though blockchain analysts later clarified that a portion of this was the thief's own funds being swapped. Beyond cryptocurrency, the attackers deployed a digital skimming module reminiscent of Magecart attacks.[1][2]
Added to the payload at 13:43 UTC, this component was designed to capture credit card numbers, expiration dates, and security codes typed directly into a website's own forms. The stolen data was then silently transmitted to an external command-and-control server.[1][4]
Targeting Mobile Devices
Fortunately, the skimmer had technical limitations that protected many consumers. It could not read payment details entered into secure, cross-origin iframes hosted by third-party processors like Stripe or Adyen, though any site processing payments natively was fully exposed.[2][3]
While the financial skimming modules cast a wide net, the OPTIMJACK campaign also included highly targeted attacks against mobile users. When the malicious JavaScript detected that a visitor was using an iPhone or an iPad, it attempted to deliver secondary payloads.[1][3]
Security firm Report URI identified these payloads as the Coruna and DarkSword iOS exploit kits. These exploit kits are designed to leverage unpatched vulnerabilities in Apple's WebKit browser engine and the underlying iOS operating system.[1][4]
If successful, they can grant an attacker deep access to the device, potentially compromising sensitive personal data and enterprise applications. The deployment of such sophisticated mobile exploits elevates the severity of the supply chain breach.[1][4]
Despite the broad distribution of the exploit kits across 178 major websites, security researchers have not yet confirmed any successful device takeovers resulting from the campaign. Devices running the latest versions of iOS 18 or iOS 26 were likely immune to the specific exploits deployed.[1][2]
The Failure of Content Security Policies
The malware also engaged in broader data collection, stealing browser cookies and capturing screenshots of the rendered pages. However, cookies protected by the HttpOnly flag remained inaccessible to the JavaScript, demonstrating the value of strict session management practices.[1][4]
The Optimizely breach serves as a stark case study in the limitations of standard web defense mechanisms. Security researchers crawled the affected organizations' websites to determine if a strict Content Security Policy would have blocked the malicious execution.[1][2]
Out of the organizations analyzed, only 15 percent had a policy strict enough to prevent the malware from running or exfiltrating data. Another 68 organizations had no Content Security Policy implemented at all, leaving their visitors entirely unprotected.[1][3]
Crucially, simply whitelisting Optimizely's domain in a security policy offered zero protection against this specific attack. Because the malicious code was appended to a legitimate script and delivered directly from the vendor, the browser's security model explicitly permitted its execution.[2][4]
Remediation and Industry Fallout
Furthermore, Optimizely's own documentation historically recommended configurations that allowed pages to run code through JavaScript's eval function. This permissive setting enabled backup copies of the malicious payload to execute on several high-profile sites, bypassing secondary defenses.[1][4]
The cleanup effort required coordinated action between Optimizely and its enterprise customer base. By 23:11 UTC on November 16, nearly all of the 188 infected snippets identified by researchers had been sanitized.[1][3]
Security experts are advising all Optimizely Web Experimentation customers to conduct thorough audits of their project configurations. Site owners should inspect their custom code fields for unauthorized base64 strings and review which API tokens possess write access to their environments.[1][2]
"The malware was a multi-purpose kit. It swapped cryptocurrency wallet addresses for the attacker's, skimmed payment cards, stole cookies and screenshots, and on iPhones and iPads it delivered the Coruna and DarkSword iOS exploit kits," noted Scott Helme, the security researcher who uncovered the extent of the breach.[1][4]
As the digital economy becomes more interconnected, the OPTIMJACK campaign demonstrates that trust is a vulnerability. Organizations must adopt defense-in-depth strategies, including protective DNS and rigorous script monitoring, to detect and neutralize malicious behavior even when it originates from a verified partner.[1][3]
Key points
- Attackers exploited a vulnerability in Optimizely Web Experimentation to inject malicious JavaScript into 178 major corporate websites.
- The OPTIMJACK malware swapped cryptocurrency wallet addresses, skimmed payment card details, and attempted to deliver iOS exploit kits.
- Because the malicious code was served from Optimizely's trusted content delivery network, it bypassed standard domain-blocking security filters.
- Optimizely has patched the vulnerability and removed the injected code, though the exact method of initial access remains undisclosed.
Open questions
- How the attackers initially gained write access across so many unrelated Optimizely customer projects.
- The total number of end users whose payment card details or cryptocurrency were successfully stolen.
- Whether any Apple devices were successfully compromised by the Coruna and DarkSword iOS exploit kits.
Timeline
Nov 16, 04:21 UTC
The first malicious injection is detected in the BBC's Optimizely snippet.
Nov 16, 13:43 UTC
Attackers update the payload to include a digital skimming module for payment cards.
Nov 16, 23:11 UTC
The vast majority of the 188 infected snippets are sanitized and cleaned.
Nov 17, 01:02 UTC
Optimizely publishes an official incident notice confirming the vulnerability and its remediation.
- Cybersecurity Researchers
- Advocate for stricter client-side monitoring and zero-trust approaches to third-party scripts.
- Enterprise SaaS Providers
- Focus on patching vulnerabilities while maintaining the flexibility of their web integration tools.
- Web Infrastructure Defenders
- Emphasize the need for defense-in-depth strategies when standard Content Security Policies fail.
Perspectives this story doesn't cover
- End users whose data or cryptocurrency was compromised
- Third-party payment processors like Stripe or Adyen
Sources
[1]Report URICybersecurity ResearchersWorld's Largest Brands Infected With Malware Via Optimizely
Read on Report URI →
[2]SecurityWeekCybersecurity ResearchersOptimizely Patched Vulnerability Exploited to Serve Malware on Major Websites
Read on SecurityWeek →
[3]Security AffairsCybersecurity ResearchersOPTIMJACK: BBC, CNN and Zoom Among Targets of Optimizely Script Hijacking
Read on Security Affairs →
[4]The RegisterWeb Infrastructure DefendersOptimizely hack turns world's biggest websites into malware dispensers
Read on The Register →
[5]OptimizelyEnterprise SaaS ProvidersOptimizely Service Status
More in Technology
See all →Supply Chain Attacks
Infoblox Discloses Corporate and Client Data Breach via Third-Party SharePoint Vendor
9 sources
AI Cyberattacks
Google Uncovers Autonomous AI Tool Converting Software Security Patches Into Active Exploits
5 sources
Zero-Day Vulnerabilities
Leaked Windows 'SignDrop' Zero-Day Exploit Grants System Control Without a Patch
8 sources
RMM Security
N-able Shuts Down N-central Remote Access Platform Following 33 Zero-Day Disclosures
6 sources
Comments
Every angle. Every day.
Get Technology stories with full source coverage and perspective breakdowns, free every day.




