Optimizely Supply Chain Attack: 178 Brands Served Malware. Here's How to Protect Your Workforce
.jpeg)
Last updated: November 17, 2026
On November 16, 2026, attackers injected malicious JavaScript into the Optimizely Web experimentation platform, and that code was served to visitors of 178 organizations, including Zoom, the BBC, CNN, Salesforce, GitLab, IKEA, HSBC, Equifax, Air France-KLM, Ledger and Peacock. The payload swapped crypto wallet addresses, skimmed payment cards, stole cookies and page screenshots, and tried to deliver iOS exploit kits to iPhones and iPads. The script itself came from Optimizely's own trusted CDN. What it fetched next did not: it called out to attacker infrastructure registered the same day. That second hop is where an on-device secure web gateway like dope.security can stop it on every employee laptop.
The original research comes from Scott Helme and the team at Report URI, who published a full breakdown and list of affected sites: World's Largest Brands Infected With Malware Via Optimizely. This post summarizes what matters for security teams and what to do about it today.
Key takeaways
What happened in the Optimizely attack?
The attacker modified the custom code inside customers' Optimizely projects, not Optimizely's CDN. Optimizely Web works by having each customer load a JavaScript snippet from Optimizely's CDN. That snippet includes an optional Project JavaScript field: custom code the customer writes in the Optimizely dashboard that runs on every page. According to Report URI, the attacker added a small loader to that field in customer projects across many unrelated organizations. The rest of each snippet was untouched, and it was still delivered from cdn.optimizely.com with Optimizely's certificate.
Report URI saw three versions of the loader: a plain script tag pointing to https://1688360.cc/track.js, the same URL hidden with base64 so a text search would miss it, and a full inline copy of the payload that ran through eval when the external domain was blocked. Sites that serve their Optimizely snippet from their own domain were affected too, because the content comes from the same project.
How the attacker got write access to so many customer projects is still unknown. Report URI says that is for Optimizely to explain, and notes that Optimizely's own sites carried the code as well.
What did the malware do?
Every visitor to an affected page received the same payload, and all of its features were active at once:
Why didn't trusting Optimizely protect anyone?
Because the malicious code arrived from the trusted domain itself. In Report URI's words, trusted JavaScript doesn't always remain trustworthy. Any control that says allow cdn.optimizely.com, whether it is a website's Content Security Policy or a corporate allowlist, let the first stage straight through.
The attack only works if the second stage succeeds. The loader had to reach 1688360.cc to fetch track.js, and the payload had to reach it again to exfiltrate cards, cookies and screenshots. That domain was registered on the morning of the attack. The iOS routers sat on .icu domains registered on November 9 and November 16. Report URI's analysis of the affected sites' own security headers makes the same point: the sites that blocked the payload were the ones that strictly limited which hosts scripts could load from and where data could be sent.
That is the lesson for everyone else. You cannot predict which trusted vendor gets compromised next. You can control where your devices are allowed to connect, and you can refuse to talk to infrastructure that did not exist yesterday.
Two different problems: your website and your workforce
This attack hit two groups, and they need different controls.
dope.security protects the first group. It is a secure web gateway that runs on your employees' Mac and Windows devices. It does not monitor or change the JavaScript your own website serves to the public, so if you run Optimizely Web, follow Report URI's site-owner guidance as well.
How does dope.security protect employees from attacks like this?
dope.security runs the secure web gateway on the endpoint, so every connection the browser makes, including the second-stage calls to attacker infrastructure, passes through policy on the device before it leaves. Here is how that applies to this campaign.

1. It blocks the infrastructure the payload depends on
The dope.endpoint performs SSL inspection and URL filtering locally, across 80+ categories. Security categories include Malicious, Phishing, Suspicious and Newly Registered, which covers domains registered in the last 30 days. Every domain this payload needed was days or hours old when it was used. With Newly Registered set to Block or Warn, a request from an employee's browser to fetch track.js or to post stolen data to card.php is stopped on the device, even though the page that triggered it came from a trusted site. Because inspection happens on the endpoint, policy applies to the full URL, not just the domain, and on any network.
2. It pushes your IOC blocklist to every laptop in seconds
When research like Report URI's lands, you need the indicators blocked everywhere, now. In dope.console, add the attacker domains to a custom category, set it to Block, and save. dope.security pushes policy to every online endpoint in seconds through push notifications, rather than waiting on a polling cycle. Devices that are offline or cannot reach the cloud keep enforcing their cached policy in Fallback Mode.
3. It tells you exactly who was exposed
Incident response starts with one question: which of our people hit this? dope.security logs each web transaction from the endpoint with the user, domain, URL path and the process that made the request. Search Analytics for 1688360.cc and the .icu hosts, or ask the Dopamine agent in plain language which users connected to those domains on November 16. You get a list of people to warn about card and crypto transactions and a list of sessions to revoke.
4. It works where your employees actually are
Employees visit CNN on hotel Wi-Fi and check HSBC from home. There is no office network to protect them. dope.security enforces the same policy on any network, in any country, with no traffic backhauled through a vendor data center.
Block these indicators now
Report URI published these attacker-owned domains. Add them, including subdomains, to a blocked custom category in dope.console or your existing controls:
See the Report URI write-up for IP addresses, file hashes and attacker wallet addresses.
What security teams should do today
FAQ
What was the Optimizely supply chain attack?
On November 16, 2026, attackers injected malicious JavaScript into the custom Project JavaScript of Optimizely Web customer projects. The code was served from Optimizely's CDN to visitors of 178 organizations and loaded a payload that swapped crypto addresses, skimmed cards, stole cookies and screenshots, and targeted iPhones with exploit kits.
Which websites were affected?
Report URI identified 178 organizations, including Zoom, the BBC, CNN, Salesforce, GitLab, IKEA, HSBC, Equifax, Air France-KLM, Ledger and Peacock. The full list is in Report URI's write-up, and the real number is likely higher.
Was Optimizely's CDN hacked?
According to Report URI, no. The CDN served exactly what it was given. The attacker had write access to customer projects, and how that access was obtained has not been explained.
Would a secure web gateway have stopped it?
For employees, a gateway that blocks newly registered and malicious domains would have stopped the payload from loading and the stolen data from leaving, because both depended on domains created days or hours earlier. Allowlisting the trusted Optimizely domain would not have helped.
Does dope.security protect my company's website from this?
No. dope.security protects your employees' Mac and Windows devices. Protecting your website's visitors requires Content Security Policy and client-side script monitoring.
Protect your people from the next one
The next compromised vendor will also be trusted. Make sure your employees' devices refuse to talk to infrastructure that did not exist yesterday. Start a free dope.security trial or book a 20-minute demo.


.jpeg)

