Researchers find Optimizely serving malicious JavaScript to sites of 178 orgs including Zoom, CNN, HSBC, Salesforce, GitLab, Ledger, BBC, Equifax, IKEA, more
First reported by Blog.report-uri ·
The trust placed in third-party JavaScript delivery platforms has been fundamentally undermined.
On November 16, 2026, a malicious JavaScript was distributed through the Optimizely Web experimentation platform, impacting 178 organizations. The injected code compromised major sites including Zoom, CNN, HSBC, Salesforce, and the BBC. This malware was capable of swapping cryptocurrency wallet addresses, skimming payment card details, stealing cookies and screenshots, and on iPhones and iPads, it delivered the Coruna and DarkSword iOS exploit kits. Verified cryptocurrency transactions show that attackers successfully stole funds. The malicious code was primarily inserted into the 'Project JavaScript' field within Optimizely's customer projects, meaning the compromised script was served from Optimizely's trusted CDN. Most of the malicious code was removed by the end of the day, with only one of the affected snippets still infected by late evening.
This incident highlights a critical vulnerability in the supply chain of web services, where a single point of compromise in a widely used platform like Optimizely can lead to widespread data theft and compromise across a vast array of high-profile clients. The attack's sophistication, including multi-stage payloads and platform-specific exploits, demonstrates the evolving threat landscape for web security. Organizations relying on such third-party tools must now re-evaluate their security postures, potentially implementing stricter content security policies (CSP) and enhanced monitoring for unexpected script behavior, even when served from trusted domains.
The broad reach of this attack, affecting sectors from finance and technology to media and retail, signals a new era of risk for digital businesses and their customers. It underscores the need for more robust third-party risk management frameworks and potentially a shift towards more defensive security measures that assume compromise rather than solely relying on trust. The financial and reputational damage inflicted, alongside the exposure of sensitive user data and the successful deployment of exploit kits, will likely spur increased scrutiny of web analytics, experimentation, and personalization platforms.
AI-written summary. May contain errors.