A vulnerability in Optimizely’s Web Experimentation account-signup process allowed an attacker to gain administrator access to customer projects and inject malicious JavaScript into websites belonging to 178 organizations. The incident included major companies such as HSBC, the BBC, Salesforce and hardware-wallet maker Ledger, according to Optimizely’s security incident report.

The attack illustrates a less visible risk in crypto payments. The blockchain networks themselves were not breached. Instead, malicious code entered through a widely used marketing and experimentation service, then reached visitors through websites that had embedded Optimizely scripts.

That distinction matters because many companies treat external scripts as marketing infrastructure rather than as part of their production software. When those scripts can execute code in a customer’s browser, however, they can influence what users see, what they copy and, in some cases, what they approve.

A trusted script became a payment threat

Optimizely said the attacker exploited a flaw in the signup flow to obtain administrative access to customer projects. The attacker then inserted JavaScript and distributed it through Optimizely’s content delivery network. The company said it fixed the vulnerability and removed the malicious code.

The injected software reportedly searched pages for cryptocurrency wallet addresses and replaced them with addresses controlled by the attacker. The affected networks included Bitcoin, Ethereum, Solana and Tron, along with five other networks. The script also modified clipboard contents, allowing a user to copy what appeared to be a legitimate address while unknowingly pasting a different one into a wallet or payment form.

Report URI researcher Scott Helme documented the malicious campaign across 178 organizations. His analysis identified crypto and trading companies among the affected businesses, as well as banks and other financial organizations. Six identified victims were crypto or trading businesses, while 32 were banks and financial companies.

The same malware was reported to include capabilities beyond address replacement. Report URI found code that could skim payment-card information, collect cookies and capture screenshots. It also identified code that served exploit kits targeting older iPhones.

Organizations affected and identified financialvictimsorganizations050100150200All affected178Crypto or trading6Banks and financial32
Organizations affected and identified financial victims

Those functions expanded the potential damage from a crypto payment attack into a broader browser compromise. A visitor did not need to hold funds on an exchange or connect a hardware wallet for the code to create risk. Simply opening an affected page could expose information or alter a transaction workflow.

Report URI-linked wallets had received approximately $7,040, mostly in bitcoin, by Monday evening, according to the investigation. That figure represents identified transfers and may not capture the campaign’s full financial impact.

Troy Hunt LM 0059 (square)
Troy Hunt LM 0059 (square) · Troy Hunt · via wikipedia · CC BY-SA 4.0

The supply-chain problem

The most important lesson is not the amount visible in the attacker’s wallets. It is the path the malware took.

Each affected organization could have maintained its own servers, authentication systems and payment infrastructure without being directly breached. The common point was a third-party service that delivered code into their sites. The arrangement allowed one account-signup vulnerability to reach unrelated businesses across sectors and geographies.

Palo Alto Networks’ Unit 42 described the incident as a software supply-chain compromise in its analysis of the attack. The researchers argued that vendor scripts with dynamic execution privileges should be treated as part of an organization’s production code, not as harmless add-ons managed only by marketing teams.

That argument challenges a common division inside companies. Security teams may monitor servers, APIs and employee devices closely while external experimentation, analytics and advertising scripts receive less scrutiny. Yet a script loaded in the browser can operate at the moment a user enters a payment amount, copies an address or submits sensitive information.

Unit 42 recommended stronger segmentation and tighter controls around vendor scripts. Its position is that organizations should limit what external code can access and isolate services that do not need to interact with payment pages or sensitive customer data.

The practical difficulty is that experimentation tools are designed to change pages dynamically. Their value comes from the ability to test layouts, messages and user flows without deploying a new version of a company’s core application. That flexibility can also make unauthorized changes harder to distinguish from legitimate ones.

Detection must extend to the browser

Sophos X-Ops independently analyzed the incident and advised organizations using Optimizely Web Experimentation to inspect their projects for injected code served on November 16, 2026, in its report on the hijacked snippets.

The recommendation points to a broader operational challenge. Traditional incident response often starts with a company’s own infrastructure. In this case, defenders also needed to examine vendor-controlled projects, identify which pages loaded the service and determine whether visitors received altered scripts.

For crypto merchants, that review should include payment pages, checkout forms and any place where wallet addresses appear. Companies also need ways to verify that an address displayed to a user matches the address ultimately sent to a transaction. Clipboard-based address replacement shows why visual inspection alone may not be enough.

The incident also reinforces the value of monitoring external JavaScript for unexpected behavior. A script that begins as an A/B testing tool may not need access to wallet fields, card data or authentication cookies. Restricting those permissions can reduce the consequences of a vendor compromise.

Crypto security has therefore become a browser and application-integrity problem as much as a wallet problem. Hardware wallets can help users protect signing keys, but they cannot by themselves guarantee that the destination shown by a compromised website is genuine. Exchanges can secure custody systems, but they cannot control every script loaded by a merchant or financial company.

Optimizely’s fix removed the immediate entry point, but the underlying business model remains widespread. Websites depend on external code to personalize content, measure campaigns and improve conversion. As crypto payments move into more commercial settings, those services will sit closer to financial decisions.

The next generation of payment security will need to protect not only private keys and blockchain transactions, but also the software that shapes what users believe they are approving.

#Optimizely#Ledger#HSBC#BBC#Salesforce#Unit 42#Sophos X-Ops#Report URI
Image credits

Jessica Jones is not a person. No notebook, no deadlines, no face behind the name — just a byline this newsroom publishes under. Here is the production line underneath it, because a name beside a portrait reads like a journalist, and this one is not one.

The models. Writing: gpt-5.6-luna. Out on the live web: gpt-5.6-luna and gpt-5.6-terra. Pictures: gpt-image-1. Swap one in the newsroom and this line swaps with it — it is read off the machines, not typed here.

How a story is made

  • Research. The searching model reads around the story, pointed at primary sources — the filing, the post, the repository — rather than at somebody else's write-up of them.
  • Writing. The writing model drafts it against what was found, at Jessica Jones's usual length and in Jessica Jones's usual register.
  • The loop. A reviewer reads the draft and sends it back with notes. Then reads it again. A piece can go round several times before it leaves the building.
  • Enrichment. A quotation has to appear word for word on the page it is taken from. A chart may only use figures that appear in the source it cites. Whatever fails is dropped, and the reason is kept.
  • Fact check. A last pass hunts for claims the article makes and its sources do not.
  • A human stop. Sensitive subjects are held for a person to read before publication, and a person can kill any of it at any point.

If that sounds less like a newsroom and more like a factory: quite. It is called Press Factory.

This article was generated using AI and published automatically without human pre-publication review.

Read and checked by admin on 11/18/2026

How this article was made

The article was produced by the Grandmonts Media News Engine using automated research, drafting and verification workflows. No human editor reviewed the article before publication. Grandmonts Media remains responsible for the published content. Errors can be reported at office@grandmonts.cz.