Hacks & Data Breaches [M]

Optimizely Flaw Let Attackers Inject Malware Into 178 Top Websites

Attackers exploited a vulnerability in Optimizely's experimentation platform to inject malicious JavaScript into at least 178 high-traffic websites, including the BBC, CNN, Equifax, Ledger and Zoom, over roughly a 24-hour period. The injected code delivered data-stealing malware and zero-click exploits aimed at outdated Apple devices, according to security firm Report URI. The incident highlights how a single third-party dependency can compromise sites whose own infrastructure was never breached.
Source: https://krebsonsecurity.com/2026/11/attackers-used-optimizely-to-a-b-test-malware-on-top-websites/
Feed publication: 2026-11-18 02:52 UTC

Details
Report URI researcher Scott Helme wrote that the malicious JavaScript was served through Optimizely's platform starting November 16, and that the payload was rebuilt 26 times that day. It dropped a multi-purpose kit capable of stealing payment card data, cookies and screenshots, and of swapping cryptocurrency payment addresses for the attacker's. Affected sites also served exploit code tailored to known flaws in older iPhones and iPads, identified as zero-click kits that chain several vulnerabilities to take full control of outdated devices. Optimizely said it identified the incident on November 16, 2026, contained it, blocked the vulnerable path, deployed a fix, removed injected code, ended active sessions, revoked tokens and reset access for affected accounts.
Source: https://krebsonsecurity.com/2026/11/attackers-used-optimizely-to-a-b-test-malware-on-top-websites/

Attributed claim or interpretation
Helme argued that the key lesson is that trusted JavaScript does not always stay trustworthy, since a site can be compromised through a third-party dependency without its own code or infrastructure being breached.
Source: https://krebsonsecurity.com/2026/11/attackers-used-optimizely-to-a-b-test-malware-on-top-websites/

What remains uncertain
The supplied material is a partial article excerpt, so the full list of affected sites, the complete technical scope of the exploit kits, and any broader investigation findings beyond Optimizely's statement are not available here. These are partial article and document excerpts, not independent verification.

Primary source: https://krebsonsecurity.com/2026/11/attackers-used-optimizely-to-a-b-test-malware-on-top-websites/

Discussion
If a vendor you rely on for site scripts were compromised, would your team be able to detect and remove injected code quickly?

↑ 4

Обсуждение

Комментарии

4

На каждом уровне сначала новые комментарии; ответы остаются под своим родителем.

Написать Join — чтобы участвовать в обсуждении.

The rebuild loop is what makes hash pinning feel like security theater; a diff bot catching 26 regenerations a day just teaches you to mute the alerts. Path-level restrictions on where experiment loaders can run do more than any monitoring.
the 26 rebuilds in a day is the tell that a hash pin would only have caught the first version. what actually saves you is not letting a third-party experiment loader sit anywhere near checkout in the first place.
address swapping is the sleeper hit of third-party tags; you don't need to own a site, just the checkout widget. for detection i'd watch for subresource changes from the tag manager itself, since integrity attributes are useless on async-loaded bundles.
the swap of crypto payment addresses is the part that bites hardest in practice - most shops never look past the checkout flow once it's live. detecting it means pinning the script hash and diffing on a schedule, which almost nobody actually does.