Optimizely Hack Pushed Malware to Sites Including CNN, BBC and Zoom
An attacker used a flaw in Optimizely's testing platform to plant crypto-stealing, card-skimming code on websites run by 178 organizations, researchers said.
An attacker exploited a flaw in Optimizely, a widely used website testing and experimentation service, to slip malicious code onto the websites of some of the best-known brands on the internet, including CNN, the BBC, Zoom and Salesforce. The code was built to steal cryptocurrency, skim payment card details and push iPhone exploits to unsuspecting visitors, security researchers said this week. Optimizely has confirmed the incident and says it has fixed the vulnerability and removed the injected code.
What happened
The malicious code first appeared early on Sunday, Nov. 16, according to an analysis by Scott Helme, founder of Report URI, a company that sells tools for detecting malicious scripts on websites. Helme examined 1,611 JavaScript files served by Optimizely and found 188 that had been tampered with, belonging to 178 organizations, BleepingComputer and the crypto news site Decrypt reported.
Optimizely’s software lets companies run A/B tests, showing different versions of a page to different visitors, by loading a small script from Optimizely’s servers onto their sites. Because the tainted files were delivered through that normal, trusted channel, the attacker did not need to break into the affected companies’ own systems. Websites simply loaded the poisoned script as they always had.
According to BleepingComputer, the injected code pulled in a second script, called track.js, from a domain the attacker controlled. In some versions, the malicious code was embedded directly in the file so it would still run if the outside script was blocked.
The organizations named in coverage of the attack include:
- News outlets CNN, Fox News and the BBC
- Technology companies Zoom, Salesforce, GitLab, Smartsheet and Bitly
- Airlines KLM and Air France, and retailers including IKEA and American Eagle
- Financial and crypto firms including HSBC’s insurance arm, the trading platforms IG and OANDA, and the hardware wallet maker Ledger
Researchers cautioned that the 178 figure counts organizations whose scripts were altered, not the number of visitors who were actually harmed. How many people loaded the malicious code remains unknown.
What the code did
The script carried several payloads at once. It scanned web pages for cryptocurrency wallet addresses on networks such as Bitcoin, Ethereum, Tron, Solana and TON and quietly replaced them with addresses controlled by the attacker, Decrypt reported, citing Report URI. It also swapped addresses that users copied to their clipboard, so a person pasting a wallet address could send funds to the wrong place without noticing.
The code also tried to capture card numbers, expiration dates and security codes typed into some sites’ own payment forms, according to BleepingComputer. Decrypt reported that card fields hosted by outside payment processors, such as Stripe, were not affected. Report URI said it could not determine how many card numbers, if any, were stolen.
On iPhones and iPads, the script attempted to load two known exploit kits, Coruna and DarkSword, which target older versions of Apple’s iOS software. Hackread reported that visitors were routed to different exploit paths depending on their iOS version, but noted that the security firm StepSecurity had not independently verified the exact exploit chains. The script also collected browser cookies, page addresses and screenshots and sent them to the attacker.
The direct financial haul appears small so far. Wallets linked to the campaign received roughly $7,000 to $7,150 in transactions, mostly in Bitcoin, Litecoin and the stablecoin USDT, according to the two outlets. BleepingComputer noted that some of those wallets were active before the attack, so the amount taken in this campaign specifically is unclear. Decrypt reported that the attacker emptied the wallets early on Nov. 17, and that a Tron address used to cash out had also appeared in earlier attacks in March and on Nov. 3.
How the companies responded
Optimizely told BleepingComputer that an unauthorized party had exploited a vulnerability and “used it to add malicious JavaScript to some customer projects.” Victoria Johnson, an Optimizely spokesperson, told Gizmodo the flaw was in “the Experimentation account signup flow,” Decrypt reported. The company has not explained how access through the signup process let the attacker modify other customers’ projects.
Optimizely said it blocked the vulnerable path, deployed a fix, removed the injected code, ended active sessions, revoked access tokens and reset access for affected accounts. It said it had contacted affected customers and would post updates on its status page. By 21:52 UTC on Nov. 16, only one of the 188 compromised files still contained malicious code, according to BleepingComputer.
Salesforce said it was investigating the modified scripts on some of its sites and had disabled Optimizely scripts as a precaution, BleepingComputer reported, and Hackread said Salesforce gave a similar account to Reuters. Salesforce said it would notify customers if it found unauthorized access to their data.
Why it matters
The incident is a textbook example of a supply-chain attack, in which hackers compromise a single trusted vendor to reach many targets at once. Modern websites routinely load code from dozens of outside providers for analytics, advertising and testing, and a breach at any one of them can spread to every site that uses it.
Helme advised people who visited affected sites on Nov. 16 to check any crypto transfers made that day, keep iPhones and iPads updated and watch card statements for unfamiliar charges. He urged website operators to review custom code in their Optimizely projects and rotate their credentials. Optimizely said its investigation is continuing.


