Hundreds of websites using the Optimizely website personalization platform served unsuspecting visitors a potent malware cocktail that targeted cryptocurrency wallets, stole payment card data, cookies, and screenshots, and delivered exploit kits for infecting iPhones and iPads, researchers said Tuesday.
ReportURI said it identified 178 websites—including CNN, Zoom, the BBC, and Salesforce—that were affected. JavaScript digitally signed by Optimizely and then loaded into customer pages had been modified to include malicious components that attempted to infect visitor devices with multiple payloads. When wares were installed successfully, they swapped user cryptocurrency wallet addresses for the attackers. Other components skimmed payment cards, stole cookies and screenshots, and, on iPhones and iPads, delivered iOS exploit kits known as Coruna and DarkSword.
Optimizely infrastructure delivers multipurpose malware
Optimizely serves JavaScript “snippets” that include both company code and code that is provided by each customer. The JavaScript, which can be served directly from Optimizely or from a customer endpoint, runs on every page the customer serves. Somehow, the attackers wrote their own code into a portion of each affected customer’s JavaScript and then delivered it from cdn.optimizely.com. A valid certificate for that domain signed the JavaScript.
ReportURI researcher Scott Helme said the delivery mechanism wasn’t a CDN cache poisoning attack, which uses HTTP to cause a CDN to cache harmful responses. Instead, the attack relied on Optimizely’s legitimate infrastructure, including its domains, to serve the tainted code. Optimizely’s own customer JavaScript for the www.optimizely.com and login.optimizely.com URLs were also affected.
“We don’t know how the attacker got write access to so many unrelated customer projects: a compromised Optimizely system, a compromised credential with cross-customer access, or something else,” ReportURI Helme wrote. “That’s for Optimizely to explain.”
Optimizely representatives didn’t respond to questions sent by email.
The scripts were named track.js, and all four components became active at the same time. Devices that received the malicious code responded differently depending on their platform and what they were doing on the final page delivering it. Actions performed included:
1. Crypto wallet address swapping. The script searches for wallet addresses for nine different cryptocurrencies and, when found, replaces them with one of the attacker’s. ReportURI said funds have already been stolen from some people whose devices ran the code.
2. Card skimming. The script watches HTTP requests for payment card numbers, expiry dates, and CVVs. When all three are present, it sends them to https://1688360.cc/card.php. Cards entered in hosted payment iframes such as Stripe and Adyen were unaffected.
3. Cookie and screenshot theft. When the script swaps an address next to a wallet or payment wording, it sends the document.cookie, the original address it replaced, and a screenshot of the page to https://1688360.cc/hit.php.
4. The DarkSword and Coruna exploit kits for iOS. The kits are delivered through a file named boot.js served from 5vobauua3ojpklx.icu and eil786npto9u5bq.icu. The file delivers a hidden iframe that contains an exploit chain tailored to the iOS version being used. Coruna can infect iOS versions 13 to 17. DarkSword can infect versions 18.x. A separate chain for iOS 26 fingerprints devices but doesn’t appear to do anything else. Earlier this month, researchers saw hackers possibly linked to the Kremlin attempting to deliver DarkSword through multiple vectors, including a compromised web analytics service.
The list of sites that ReportURI found delivering the malicious scripts numbers 178, and as noted, includes many big names. It’s unclear if the list is exhaustive. Helme said in an interview that “Ars wasn’t affected as far as we saw, and I haven’t seen any other indications it was.” Website administrators can check both the ReportURI post and this one from security firm Cynet for indicators of compromise.
At the moment, there’s no information available on how successful the exploit chains were in installing the payloads. Beyond the bare-bones details above, more precise information about what versions of iOS were susceptible was also unavailable. These sorts of gaps are common in breaking news concerning mass malware attacks, because it takes time for researchers to sift through the evidence. Readers are encouraged to check back as new developments come to light.
The indicators may help end users determine if they are affected, but the skill required is likely out of reach for many.

