1. Optimizely Hack Turns 178 Organizations' Scripts Against Their Visitors
Websites belonging to 178 organizations, including hardware wallet maker Ledger, Salesforce and the BBC, delivered malware to their own visitors on Monday, November 16, 2026, after attackers planted code in scripts hosted by the website optimization platform Optimizely. The Optimizely hack was built to swap cryptocurrency wallet addresses and capture payment card details typed into web forms, security researchers said.
The count of affected organizations comes from an investigation by Report URI, a web security monitoring company led by chief executive Scott Helme. Researchers tracking the campaign have labelled it Optimjack. Optimizely has confirmed that an unauthorized party abused a weakness in one of its account processes to insert malicious JavaScript into some customers' projects, and says it has since closed the hole and stripped out the injected code.
The financial damage identified so far is small when set against the reach of the attack. Wallet addresses tied to the malware took in roughly $7,000 in cryptocurrency on Monday. What cannot yet be measured is how many card numbers, and how much other browser data, reached the attackers before the code was removed.
2. How did the Optimizely hack work?
The attackers did not break into Ledger, the BBC or any other victim directly. They altered code that those organizations had already chosen to load from Optimizely, so each affected site delivered the malware itself, through a script it trusted.
Optimizely sells tools that let businesses test variations of their web pages and run custom code for experiments. To use the service, a company places a JavaScript snippet on its site. That snippet pulls in the customer's project from Optimizely, including any custom code the customer has written for it, and executes it in every visitor's browser. The design is what makes the product useful, since marketing and product teams can change a live site without a full software release. It is also what made the attack effective.
By inserting malicious instructions into customer projects, the intruders ensured that an ordinary, expected script carried the payload whenever someone opened an affected page. Nothing looked out of place to the visitor. The address bar showed the genuine site, the connection was encrypted as usual, and the page behaved normally while the added code ran alongside it.
This pattern is known as a supply chain attack. Instead of compromising many targets one at a time, an attacker compromises a single supplier whose code those targets all run. Third-party scripts are a particularly attractive route because they execute with the same privileges as the host site's own code and can read or change almost anything on the page.
3. Which Websites Were Affected
Report URI's review found tampered script assets belonging to 178 organizations. In the crypto and trading sector the list includes Ledger, one of the largest makers of hardware wallets for storing digital assets, along with OANDA and its crypto arm Coinpass, and the trading platforms IG, Interactive Investor and TMS Brokers.
Beyond finance, the affected names include IKEA, Salesforce, Zoom, the BBC and CNN. Together these properties receive millions of visits a day, which gave the attackers an audience far larger than a typical phishing operation could assemble.
The presence of financial services firms on the list is significant because of what the malware was built to do. A visitor who arrives at a wallet maker's site or a brokerage is more likely than the average web user to be copying a deposit address or entering payment information. Ledger, Coinpass and other organizations named by researchers were approached for comment, and no responses from them have been reported.
Helme said the BBC was among the first sites his company observed serving the injected code, in the early hours of Monday in UTC terms. The bulk of the malicious code had been cleared out by 9:52 pm UTC the same day, meaning the campaign was live across many high-traffic sites for much of a working day in Europe and the United States.
4. Optimizely's Account of the Breach
A spokesperson for Optimizely said on Monday that an unauthorized party exploited a flaw in the signup flow for the company's Experimentation accounts and used it to add malicious JavaScript to some customer projects. "We have blocked access to the vulnerable path and deployed a fix," the spokesperson said.
The company described a series of containment steps. It removed the injected code from the projects concerned, terminated active sessions, revoked tokens and reset access for the affected accounts. It also blocked delivery of the code through Cloudflare, the network that sits in front of its content. Optimizely said its investigation is continuing and that every affected customer has been notified.
Researchers examining the campaign from the outside said they had not established how the attackers obtained the ability to modify Optimizely projects. Their analyses were based on what the scripts did once they reached browsers, not on visibility into Optimizely's internal systems, so the company's explanation is the only account so far of the point of entry.
5. What the Malware Could Steal
Three capabilities stand out in the researchers' descriptions of the code. The first targets cryptocurrency transfers. Once running in a visitor's browser, the script could replace legitimate wallet addresses with addresses controlled by the attackers, including an address the visitor had copied to the clipboard. Because crypto addresses are long strings of characters that few people check in full, a user could paste the substituted address into a transfer and send funds to the wrong destination without noticing. Blockchain transactions cannot be reversed once confirmed.
The second capability is card skimming. The script monitored form fields for card numbers, expiry dates and security codes, then transmitted what it found to a server run by the attackers. This technique has been used against online shops for years. Its appearance on brokerage and media sites reflects how broadly a compromised third-party script can be repurposed.
The third applies to Apple devices. According to Helme, people who visited the affected sites on iPhones and iPads faced an additional risk, because the script loaded code associated with two iOS exploit kits known as Coruna and DarkSword. Exploit kits attempt to take advantage of weaknesses in a device's software, which goes beyond stealing what a user types into a page.
Analysis by the cybersecurity firm StepSecurity examined the same campaign and reached matching conclusions about the address-swapping and card-capturing functions.
6. How much crypto was stolen?
Confirmed crypto losses are low: addresses linked to the malware received roughly $7,000 in cryptocurrency on Monday, according to blockchain records. That figure covers funds arriving at the attackers' known addresses and does not prove that every payment came from a victim of this campaign.
The larger unknown is the card data. Helme said he could not determine how many payment card details or browser cookies reached the attackers' server. Stolen card numbers are usually sold or used later, so any resulting fraud would not appear in the on-chain figures and may take weeks to surface in bank statements.
The gap between the scale of exposure and the sums recovered on-chain is typical of clipboard-swapping attacks. They succeed only when a visitor happens to copy and paste a crypto address during the window in which the code is live, and only if that person fails to check the pasted result. Millions of page views therefore translate into a small number of misdirected transfers.
7. What Users and Site Operators Should Take From It
For individuals, the incident reinforces a habit that security professionals have long recommended for crypto transfers: confirm the full destination address after pasting it, and verify it on a hardware wallet's own screen where one is used, since that display is not controlled by the web page. A genuine domain name and a valid padlock icon were no protection on Monday, because the malicious code was delivered by the legitimate sites.
Anyone who typed card details into one of the affected sites on Monday has reason to watch statements closely and to contact the card issuer if unfamiliar charges appear. The same applies to people who moved crypto after copying an address on one of those pages: the transaction history will show whether funds arrived where they were meant to go.
For companies, the episode is a reminder that every third-party script on a page is part of that page's attack surface. A business that embeds an outside vendor's code is extending its visitors' trust to that vendor's security practices, including how the vendor controls who can create accounts and edit projects.
Experimentation and analytics tools are among the most widely deployed scripts on the commercial web, and they are often added by teams outside the security function. Monday's events show how a weakness at one such provider can reach wallet makers, brokers, broadcasters and retailers at the same moment, regardless of how well each of those organizations protects its own servers.
8. What Comes Next
Optimizely's investigation remains open, and the company has not yet published a full account of how the signup weakness was used to reach customer projects. That explanation, when it arrives, will show whether the attackers needed anything more than the flaw the company has described.
Responses from the affected organizations are also outstanding. Ledger, Coinpass and the other named firms have yet to say publicly whether their own reviews found evidence that customers were harmed. Until those findings and any reports of card fraud emerge, the roughly $7,000 traced on-chain is the only confirmed measure of what the campaign took.
