Italy's ACN warned on 2 November about four flaws rated 8.5 under CVSS 4.0. Three in FTS5 and the session extension were fixed in 3.54.0; the .recover command injection is listed through 3.54.1 with no fixed release.
Attacker injected malicious JavaScript into Optimizely customer projects
EditorialBy TrustList Editorial
The experimentation vendor says it has fixed the flaw and removed the code; a researcher counted 188 infected scripts, a figure Optimizely has not confirmed.
- Cybersecurity
- Application Security
- A/B Testing
- Web Analytics
- +1 more
About Attacker injected malicious JavaScript into Optimizely customer projects
Attacker injected malicious JavaScript into Optimizely customer projects
17 November 2026: Optimizely says an unauthorised party exploited a vulnerability in its Web Experimentation product and used it to add malicious JavaScript to some customers' projects, code that would then have run in the browsers of visitors to those customers' websites. The company says it has fixed the flaw, which has no CVE number, and removed the injected code; it has not said how many customers were affected or for how long.
Not yet independently verified. Optimizely's status notice of 17 November 2026 does not name the vulnerability, the number of customers affected or the period of exposure, and gives customers no instructions. The counts of infected files and organisations, and the description of what the injected code did, come from a researcher quoted by BleepingComputer; we have not read the researcher's own write-up. We will update this when it can be confirmed, and remove this note.
What Optimizely has confirmed
The statement is an entry on the company's status page, opened early on 17 November UTC against the Snippet Delivery component of Web Experimentation and marked as being monitored. It reads: "Earlier today, Optimizely identified a security incident in which an unauthorized party exploited a vulnerability in Experimentation and used it to add malicious JavaScript to some customer projects. We've fixed the vulnerability and removed the injected code from affected projects. We've also taken steps to secure affected user accounts and put additional protections in place".
That is the whole of the vendor's account so far. It gives no technical description, no list of affected customers and no advice.
What a researcher reports
BleepingComputer reports the findings of the security researcher Scott Helme, who examined 1,611 Optimizely scripts on 16 November and found 188 of them infected, belonging to 178 organisations. By his account the injected code loaded a second script from an outside server that could swap cryptocurrency addresses, skim payment card details, and take cookies, page addresses and screenshots. He found one of the 188 files still infected late on 16 November.
These figures are the researcher's, relayed by the press. Optimizely had not answered the outlet's questions when it published, and it is not yet clear how one flaw gave access to so many separate customer projects.
Why an experimentation snippet is a high-value target
An A/B testing tool works by loading a script from the vendor on every page under test, and that script is designed to rewrite the page. Whoever controls it controls what the visitor sees and types. That makes the snippet as sensitive as the site's own code, although it is usually added by a marketing team and reviewed by nobody in security.
Checks for a site that loads the snippet
Optimizely's notice asks nothing of customers, so the following is drawn from the nature of the incident, not from the vendor.
A site owner can compare the project's current custom code and experiment code in the Optimizely console against what the team wrote, and review the project's change history and user list for entries nobody recognises. Web server and content security policy reports for 16 November will show whether browsers were asked to load scripts from an unfamiliar host. If the site takes payments on pages that carry the snippet, the payment provider and the card scheme rules on reporting apply, and the question to put to Optimizely in writing is whether the project was among those altered.
A longer-term control is a content security policy that limits which hosts a page may load scripts from, which would have stopped the second-stage script.
Our A/B testing software category lists 11 products that work the same way and carry the same exposure. Optimizely's status entry remains open and is the place where an update will appear.
Sources
- Optimizely status: Web Experimentation Security Incident, 17 November 2026
- BleepingComputer: Optimizely supply chain attack infects major brands' sites with malware, 17 November 2026
Categories & features
- Cybersecurity
- Application Security
- A/B Testing
- Web Analytics
- Data Security
TrustList Weekly
The week in software and IT, in one email
The news that matters to buyers, new rankings and our own research. Every Thursday, free, and easy to leave.
You may be interested in
An attacker swapped one customer's page-speed script on 29 October and used it to drain trading accounts on Polymarket, which is now returning $3.9 million to traders.
Khosla Ventures and Hanabi Capital co-led the round, taking the three-year-old San Francisco company to $35m in total, with Dropbox and Block among named customers.
The French regulator's final recommendation covers the vendors that build replay tools and the sites that run them. No exemption applies, purposes must be set before collection, and mouse movement should be masked where login codes are typed.
Two weeks after a tampered coding extension appeared in its marketplace, Cursor says at most 5,000 users could have received it, and that it has moved all inference away from the vendor where the attack began.
NHN Dooray says attackers passed its password reset by repeating codes and used API keys leaked through an Open API flaw. It reported to KISA on 13 November and to the PIPC on 14 November.
A Hamburg security firm found records with email addresses in 7,403 of them, from passport numbers to chat logs, and says the 30% share is not a confirmed leak rate.
The New York and Tel Aviv data security company says its revenue grew more than 300% in a year. The money goes to controls over what AI agents, copilots and models can reach, and to sales in North America and Europe.
More on TrustList
Everything here links back to the same verified catalogue. Pick your next stop.
- More CybersecurityThe ranking for this subject
- CompaniesAgencies, consultancies and IT service providers, ranked by verified reviews.
- ProductsSoftware and SaaS with pricing, features, integrations and alternatives.
- AwardsAnnual recognition decided by verified reviews and an independent jury.
- LaunchesNew products and releases, voted up by the community every day.
- AI ModelsBenchmark scores and community ratings for every major model.
- RequestsBuyers describe what they need; vendors respond directly.
- PeopleReviewers, authors and makers with public profiles.
- ComparePut up to four listings side by side before you shortlist.