Optimizely supply chain attack

The Optimizely Supply Chain Compromise: Card Skimmers on Trusted Sites

On November 17, 2026, security researchers discovered that attackers had exploited a vulnerability in Optimizely Web Experimentation to inject malicious code directly into the JavaScript files that major websites—CNN, BBC, Zoom, and at least 176 others—load into visitor browsers. Instead of protecting your data on trusted sites, you may have been silently served card-stealing malware and mobile device exploits. This is a supply chain attack: the attacker compromised the middleman, not the destination.

Optimizely Supply Chain Attack: What Happened and What You Need to

How the Attack Worked

Optimizely Web Experimentation is a tool that websites use to test different versions of pages and personalize user experience. It works by loading a JavaScript file from Optimizely's servers into every page that uses it. An attacker found a way to compromise this file delivery mechanism and inject malicious code into the scripts that Optimizely served to customer websites.

When a visitor arrived at an affected site—say, CNN or BBC—their browser downloaded what appeared to be the legitimate Optimizely script. Instead, it contained hidden card-skimming code designed to harvest payment card details typed into forms, and iOS exploit kits aimed at compromising mobile devices. The visitor had no way to detect this. The site's owner did not intentionally serve malware. Report URI, a Content Security Policy and script monitoring firm, first detected the malicious payloads across 178 organizations and alerted Optimizely on November 17.

Why Supply Chain Attacks Are So Effective

A supply chain attack works because it breaks the trust relationship between a website and its visitors. A visitor trusts CNN or the BBC because of their reputation. They trust that these sites would not knowingly serve malware. That trust is then weaponized: the attacker does not need to breach CNN's servers; instead, they compromise a service that CNN relies on, turning CNN's own trust into a backdoor.

Optimizely serves thousands of websites, so compromising its script delivery system exposes millions of end users in a single action. The attacker does not need individual access to each site. This is why supply chain vulnerabilities often affect large numbers of targets at once, and why they are extremely difficult for individual website operators to prevent. The website owner cannot verify every third-party script it loads; they depend on the third-party vendor to secure it.

What the Malicious Code Actually Did

The injected JavaScript performed two primary functions. The first was card skimming: monitoring form inputs and stealing payment card data, expiration dates, and security codes as users typed them into checkout pages or donation forms. This data would then be exfiltrated to attacker-controlled servers. The second component was an iOS exploit kit, meaning it attempted to detect iOS devices visiting the affected pages and deliver targeted vulnerability exploits to compromise them.

The iOS exploit approach is notable because it suggests the attacker was willing to invest in multiple attack vectors. Card skimmers are broadly used and relatively reliable at scale. iOS exploits, by contrast, are far more expensive and are typically reserved for high-value targets or state-sponsored operations. The presence of both tools in one campaign hints that the attacker either had funding or was testing different payloads for future use.

Context: How This Fits Into Modern Cybersecurity Risks

Third-party JavaScript vulnerabilities are well-documented by security researchers. A compromise of a single widely used script can affect thousands of sites without the need for individual breaches. Tor Project documentation on onion service vulnerabilities reminds us that the attack surface extends beyond direct compromises to include trusted infrastructure. Public law-enforcement statements on supply chain incidents (from the CISA and FBI) emphasize that detection often comes from security monitoring firms like Report URI rather than from the affected organizations themselves. This matters because it means compromises may persist undetected for days or weeks. Court records from previous supply chain cases show that attackers often time their injections to maximize profit before detection, then clean up the malicious code once discovered. Understanding this pattern helps explain why immediate patching and forensic investigation are critical after these attacks are disclosed.

What You Should Do Now

If you visited CNN, BBC, Zoom, or any other major website between November 16 and November 17, 2026, and entered payment card data or used iOS, consider the following steps:

  1. Monitor your credit card statements closely for unauthorized transactions over the next 90 days.
  2. If you entered a card number, contact your card issuer directly (use the number on the back of your card, not a number from an email) and ask them to review your account for fraud.
  3. If you use iOS and visited one of the affected sites, install the latest security updates from Apple immediately.
  4. Use a password manager to change any passwords you entered on these sites, and ensure each is unique.
  5. If you are concerned about session compromise, log out of all sessions on those websites and clear your browser cache.
  6. Monitor your Apple ID for unauthorized sign-in attempts if you have an iPhone or iPad.

Website owners and developers should review the list of affected organizations published by Report URI and confirm whether their own sites were compromised. If you manage a website that uses Optimizely, contact their support team for guidance on whether your instance was affected and what remediation steps to take.

Lessons for Site Operators and Users

This incident illustrates why relying on a single third-party vendor creates cascading risk. For website operators, the lesson is to audit third-party dependencies, use Subresource Integrity (SRI) hashing to verify that scripts have not been tampered with, and implement strong Content Security Policy headers that restrict what injected scripts can do. For ordinary users, the lesson is humbler: even reputable websites can unknowingly serve compromised code, and no amount of vigilance will always prevent exposure. What you can control is how you respond once a compromise is public.

The Optimizely incident also reflects a broader pattern: attackers are increasingly targeting the ecosystem of tools and services that power the internet, not just end users. This means security vendors, hosting providers, and infrastructure companies are now front-line targets. Monitoring your own internet activity—using browser extensions that check for known malicious scripts, keeping your operating system and applications fully patched, and using a reputable VPN when on public networks—remains practical steps, though they are not panaceas.