Browser Booby-Trap Nails CNN, BBC, Zoom

Hooded figure using a laptop with digital code overlay
Photo: metamorworks / Shutterstock

Attackers slipped card-stealing and crypto-swapping code onto 178 organizations’ sites through a single trusted service, and most users never had a chance to see it coming.

Story Highlights

  • Optimizely confirmed a vulnerability let an outsider add malicious JavaScript to some customer projects.
  • Independent scans found 188 tainted files linked to 178 organizations, including CNN, BBC, Fox News, and Zoom.
  • The code acted like a card skimmer and a crypto “clipper,” pointing to a supply-chain style attack.
  • Optimizely blocked the path, removed injected code, and reset access, but losses and attribution remain unclear.

What Optimizely Said Happened

Optimizely said an unauthorized party exploited a flaw in its Experimentation account signup flow. Using that access, the attacker added malicious JavaScript to some customer projects. Optimizely reported it blocked the vulnerable path, deployed a fix, removed the injected code, ended active sessions, revoked tokens, and reset access for affected accounts. The company framed the impact as affecting some Web Experimentation customers and said monitoring and investigation would continue.

Researchers and reporters tied the injected code to customer project JavaScript that loads on every page. That matters because many sites trust Optimizely’s scripts to run experiments. If that shared layer is poisoned, a website can be hit even if its own servers are clean. This is a classic client-side supply-chain problem. The Optimizely content delivery network delivered the altered snippets, which then ran in visitors’ browsers on those branded sites.

How Big the Blast Radius Was

Independent scanning backed up a wide scope. StepSecurity reported 188 infected Optimizely JavaScript files across 178 organizations after examining 1,611 scripts. The list included high-profile brands like CNN, BBC, Fox News, Zoom, Salesforce, GitLab, HSBC, KLM, IKEA, and Ledger. These names show reach, not guaranteed theft per site. The count reflects poisoned snippets tied to those organizations, not a confirmed log of every visitor exposure or loss at each domain.

The malicious functions fit a clear pattern. StepSecurity said the payload loaded two tools: a crypto “clipper” that swaps wallet addresses, and a card “skimmer” that tries to collect payment data from web forms. Cointelegraph summarized the same behaviors. This mix targets two growing parts of the economy at once: online checkouts and digital assets. It also aligns with past client-side attacks that ride third-party code to grab money at the edge, inside the browser.

What We Still Do Not Know

Key facts remain open. Optimizely has not shared the exact technical flaw, the exploit steps, or whether attackers needed to create or use certain accounts. No public tally shows how many users lost money, which sites saw actual skimming runs, or how much cryptocurrency moved due to address swaps. Investigators have not named the group behind the campaign. Public sources have not published the full dataset of infected code or a site-by-site execution record.

These gaps highlight how the public learns about cyber events. Vendors control internal logs and tokens. Customers worry about brand risk and lawsuits. Reporters chase big names because readers know them. That can blur lines between “infected snippet,” “executed code,” and “confirmed theft.” Still, Optimizely’s own admissions and the consistent technical findings support the core facts: a shared service became a single point of failure for many well-known sites, at least for a time.

Why This Matters Beyond Tech Circles

Everyday people used trusted sites and still faced hidden risk. They did nothing wrong. Companies outsourced a piece of their pages and paid for it with trust. This is the deeper worry many Americans share about big systems run by far-off players. A small flaw in a vendor can ripple across news outlets, airlines, banks, and stores in hours. When accountability is spread thin, regular users and small businesses carry the cost first.

There is a practical lesson both sides can back. Websites should cut unneeded third-party scripts, lock down what remains, and watch what runs in the browser. Content Security Policy, subresource integrity, and real-time reporting can block or flag rogue code. Vendors should publish full incident reports fast and prove cleanup with data. Regulators should set clear rules for notice and testing. These steps will not end attacks, but they can shrink the window when bad code runs free.

Sources:

insiderpaper.com, thenextweb.com, cointelegraph.com, infostealers.com, support.optimizely.com