Key takeaways for the Optimizely cyber attack
- Attackers exploited a flaw (CVE-2026-132224) in Optimizely's Web Experimentation product and planted malicious JavaScript in customer projects on November 16, 2026.
- Researchers found infected Optimizely snippets associated with 178 organizations, including major brands and Optimizely's own websites.
- The compromised snippets were served from Optimizely's CDN or, in some cases, customer domains. They loaded malicious code from attacker-controlled infrastructure, with an inline fallback in some versions.
- Optimizely says it fixed the flaw, removed the injected code, ended active sessions, revoked tokens, and reset access for affected accounts. Researchers reported card skimming and cryptocurrency theft.
- Any team that runs Optimizely Web Experimentation should check the Project JavaScript field in every project for code it did not write.
On November 16, 2026, attackers used a vulnerability, CVE-2026-132224, in Optimizely's Web Experimentation platform to inject malicious JavaScript into customer projects. The attack was dubbed ‘The OPTIMJACK’ by security researchers. Researchers were able to identify infected snippets associated with 178 organizations after examining 1,611 Optimizely snippets. The snippets were served from Optimizely’s CDN or, in some cases, from customer domains; they then loaded the attacker’s code or ran an inline fallback. Optimizely confirmed the incident on November 17. The attack is a reminder that a vendor you trust can become the path malware takes into your visitors’ browsers.
What happened?
On November 17, Optimizely posted an incident notice saying an attacker exploited CVE-2026-132224 in Experimentation and used it to add malicious JavaScript to some customer projects. Optimizely told BleepingComputer that the flaw was in the Experimentation account signup flow. The malicious code was added to a field called Project JavaScript, which runs custom code on pages where the snippet loads. BleepingComputer reported that a researcher found 188 infected script files across 178 organizations including Optimizely's own website and login portal. Beazley Security reported related malicious scripts involving three other services, with earlier activity observed from November 13.
Why did this attack work so well?
The compromised snippets still appeared to come from Optimizely’s CDN, over a connection using Optimizely’s certificate, and the attacker left the rest of each snippet intact. That made the snippet itself appear to come from a trusted source. The injected loader usually fetched the payload from an attacker-controlled domain, which could be visible in browser or network monitoring. Some versions also included an inline fallback that ran if the external payload was blocked.
Optimizely’s Content Security Policy documentation says some features require unsafe-eval in the script-src directive. Security researchers reported that this setting allowed an inline fallback copy of the malware to run on a company’s website. Teams using Optimizely should review whether their Content Security Policy permits unsafe-eval, whether their configuration requires it, and whether they can remove it.
Researchers reported that the malware could swap cryptocurrency addresses and skim payment-card details. Related reporting also described code that attempted to deliver iOS exploits. Hudson Rock said the available evidence did not establish successful device exploitation. Researchers identified Coruna and DarkSword in the exploit chain related to iOS, both are well-known iOS exploit kits and frameworks.
Who is affected?
Any organization that uses Optimizely Web Experimentation, and any visitor to an affected site, may have been exposed. Earlier this year, Optimizely disclosed a separate vishing incident that exposed basic business contact information.