Optimizely Attack Reached 178 Organizations: What Website Owners and Visitors Should Check

Malicious Optimizely scripts reached sites belonging to 178 organizations, researchers found. Here is how the attack worked and what operators and visitors can check.
A code editor displaying server-side JavaScript
Photo by James Wiseman on Unsplash.

Attackers exploited a vulnerability in Optimizely’s website experimentation platform to insert malicious JavaScript into customer projects, exposing visitors to code designed to steal payment details and divert cryptocurrency transfers. Optimizely says it has fixed the vulnerability, removed the injected code and notified affected customers.

Security researcher Scott Helme’s team at Report URI found 188 infected scripts belonging to 178 organizations among 1,611 snippets it examined on November 16. The affected organizations included Zoom, the BBC, CNN, Salesforce and GitLab. That is a count from a sample, not a total of all affected customers or compromised visitors. Report URI’s investigation and BleepingComputer’s independent checks documented malicious scripts on a number of those sites.

Being listed does not mean every visitor had data stolen or every page on a company’s site was affected. The number of cards, cookies and devices actually compromised remains unknown.

How a trusted script carried the attack

Optimizely Web Experimentation helps site owners test changes by loading a JavaScript snippet into their pages. According to Report URI, the attacker placed code in the customer project’s custom JavaScript field. The resulting snippet still came from Optimizely’s normal content delivery network, so an organization’s own web server could be clean while its visitors received the injected code.

One version loaded another script from 1688360[.]cc. Researchers also found an encoded loader and an inline variant, so a simple search for that domain alone is not enough to rule out exposure. Some customers served the same project content through their own domains.

Optimizely told BleepingComputer that the flaw was in the Experimentation account signup flow. It reported blocking the vulnerable path, ending active sessions, revoking tokens and resetting access for affected accounts. In its service update, the company described the vulnerability as fixed and the injected code as removed. Its investigation continues; a complete public account of the intrusion and its duration has not been established.

What the malicious code tried to do

Report URI found a cryptocurrency address swapper that could replace an intended wallet address in page content or the clipboard with an attacker’s address. The team also identified a card skimmer that watched for card details submitted through a site’s own forms. Its analysis found that separately hosted payment fields in third-party iframes were outside that particular skimmer’s reach, though operators should verify their own implementation.

The researchers also observed code that could collect accessible cookies and screenshots and route some iPhone and iPad visitors toward exploit kits. Delivery of an exploit is not proof that a device was successfully compromised. Report URI could not determine how many devices were infected or how many payment cards were captured.

Steps for site operators

  1. Establish exposure. Inventory pages and projects that load Optimizely Web Experimentation, including snippets served through your own domain. Contact Optimizely and preserve relevant project history, logs and copies of the scripts before changing evidence.
  2. Inspect what was actually served. Compare downloaded snippets and project and experiment code with approved versions. Search for the indicators in Report URI’s technical report, but investigate unknown loaders and encoded or inline code too. A clean current file does not establish what visitors received earlier.
  3. Contain and assess. Review whether to disable affected snippets, especially on login or payment pages, while investigating. Review user, integration and API access; rotate credentials and potentially exposed session tokens in coordination with your incident response team.
  4. Investigate payment exposure. Identify whether card fields were on your own page or isolated in a hosted payment iframe. If payment data may have been exposed, involve your payment provider and incident response and legal teams promptly.
  5. Review browser defenses. Examine Content Security Policy reports for unexpected script and connection destinations. A restrictive policy can block some external loads, but it is not proof of safety against every inline or allowed-host variant.

Security firm Beazley Security separately documented malicious code associated with scripts from other providers. Its advisory predates Optimizely’s remediation statement, but its technical indicators and advice to inspect the actual delivered files remain useful for an investigation.

What visitors can do

If you used a potentially affected site on November 16, check its own incident notices for details. Review payment-card activity if you entered card details on an affected page, and contact your card issuer promptly about unfamiliar transactions. If you sent cryptocurrency, compare the destination recorded in the transaction with the intended address; a completed transfer generally cannot simply be reversed.

Install the latest security updates available for your iPhone or iPad. Updating is a sensible precaution, but it cannot by itself establish whether an earlier visit led to compromise. People who suspect account misuse should change affected passwords, revoke sessions where available and enable multifactor authentication.

Leave a Reply

Your email address will not be published. Required fields are marked *

Previous Post
Rows of server racks inside a modern data center, used to illustrate AI infrastructure capacity

U.S. AI Chip Rule Targets Remote Cloud Access: What Providers Must Check

Related Posts