Optimizely Script Hijack: What to Check If Your Website Uses Optimizely
On November 16, 2026, attackers planted a card skimmer and a crypto wallet swapper inside Optimizely website testing scripts used by 178 organizations. If your website uses Optimizely Web, your visitors may have run that code. Optimizely says the code has been removed, but only your own records can show whether your site was affected.
Check your site in under a minute. Our free website check now looks for this attack. It tells you whether your site loads Optimizely, whether the script contains the attacker's code today, and whether your security policy would have blocked it.
What happened?
Optimizely Web is a testing service that lets companies run A/B tests on their websites. To use it, a company adds one Optimizely script to its pages. That script is hosted on Optimizely's servers and includes a "Project JavaScript" field where the company can add its own code.
On November 16, an unauthorized party exploited a flaw in Optimizely's Experimentation product and added malicious code to the Project JavaScript of many customers' projects. Because the code was delivered inside each company's own Optimizely script, from Optimizely's servers, every visitor's browser treated it as trusted.
Researcher Scott Helme of Report URI found the attack. His team checked 1,611 Optimizely scripts and found the malicious code in 188 of them, belonging to 178 organizations. (Report URI)
Timeline (November 16, 2026, UTC)
| Time | What happened |
|---|---|
| About 04:20 | First injected Optimizely script seen |
| 08:10 | First version of the crypto wallet swapper |
| 13:43 | Card skimming added |
| 17:13 to 17:57 | Attacks on older iPhones expanded |
| 20:40 to 21:10 | Mass cleanup of the injected code |
| 23:11 | All infected scripts observed clean |
The attacker's main domains were suspended early on November 17. Source: Report URI, with technical detail from StepSecurity.
Who was affected?
Report URI identified 178 organizations. The largest groups were banking and finance (32 organizations), technology and software (23) and retail and fashion (20). Optimizely's own websites were among those affected. (Report URI, Help Net Security)
Optimizely has acknowledged the incident. According to StepSecurity, Optimizely says it fixed the flaw, removed the injected code from affected projects, secured affected accounts and added protections against unauthorized code delivery. As of November 17, Optimizely had not published how the attacker gained access to so many customers' projects.
What did the malicious code do?
- Swapped crypto wallet addresses. When a visitor copied or pasted a wallet address near words such as "send" or "deposit," the code replaced it with one of the attacker's addresses.
- Stole payment card details. It watched forms and network requests on the page for card numbers, expiry dates and security codes, and sent them to the attacker. Card fields hosted by a payment provider in a separate frame were out of its reach.
- Took screenshots and cookies. It captured an image of the page and any cookies that were not marked HttpOnly.
- Targeted older iPhones. On iPhones and iPads running older versions of iOS, it loaded further attack code from other servers.
The attacker changed the code 26 times during the day, so defenders should look for behavior and known addresses rather than a single file fingerprint. (StepSecurity)
Why didn't normal security tools catch it?
Nothing changed on the affected companies' own servers. File integrity monitoring, server malware scanners and most web application firewalls watch your server, so they had nothing to see. The code arrived from a trusted, long-approved vendor, over a valid HTTPS connection, inside a script the company had chosen to load.
This is a third-party script supply chain attack. Report URI also links the same actor to earlier injections through other third-party scripts in the days before, so the lesson applies to every outside script on your site, not only Optimizely.
How to tell if your website was affected
- Run a quick outside check. Our free check finds your Optimizely script and scans it for the attacker's known addresses and code. A clean result today is good news, but it does not prove your site was clean on November 16, because Optimizely has since removed the code.
- Review every Optimizely Web project. Look at the Project JavaScript and any experiment custom code for
1688360[.]cc,webshell[.]lol,atob(and scripts you do not recognize. Encoded copies will not show up in a plain text search, so read the code, not just search it. - Check change history and access. Review each project's change history for edits on or around November 16, and remove users, API tokens and integrations you do not recognize.
- Ask Optimizely directly. Ask whether your projects were modified, when, and through which account.
- Search your logs. Look in web proxy, DNS and security tool logs from November 13 onward for the domains
1688360[.]ccandwebshell[.]lol, and for the other addresses in the indicators of compromise below. If you collect Content Security Policy reports, look for blocked requests to them too.
If your site was affected: what to do
- Remove the code from every affected Optimizely project if any remains, and rotate Optimizely credentials and API tokens.
- Treat card details as exposed if customers entered them in forms on your own pages on November 16. Your payment processor can advise on next steps.
- Rotate secrets in readable cookies. Any session token or secret stored in a cookie that is not marked HttpOnly may have been captured.
- Talk to legal counsel early about whether customer notification is required, and preserve your logs and Optimizely change history as evidence.
- Get help if you need it. Our incident response team is available 24/7 at 1-800-390-1053.
What would have stopped it?
A Content Security Policy (CSP) is a header that tells browsers which addresses your pages may load scripts from and send data to. A strict policy would have blocked the attacker's script from loading and blocked the stolen data from leaving the page, even though the loader came from a trusted vendor.
Very few affected sites had one. Of the affected sites Report URI could evaluate, 68 had no policy at all, and only 24, about 15%, had a policy that blocked the attack. (Report URI) Two common settings let it through:
- 'strict-dynamic' lets a trusted script, such as the Optimizely script, load any other script.
- 'unsafe-eval' allowed one version of the attack to run a copy of its code placed directly in the Optimizely script.
A practical policy for most sites lists approved script sources in script-src, lists approved destinations in connect-src, avoids 'unsafe-eval', and turns on reporting so you learn when something is blocked.
Payment pages have a compliance angle
If you take card payments, PCI DSS version 4.0.1 already expects this kind of control. Requirement 6.4.3 calls for an inventory of every script on payment pages, with each one authorized and its integrity checked. Requirement 11.6.1 calls for a way to detect unauthorized changes to payment pages. Both became mandatory on March 31, 2025. (PCI Security Standards Council) Moving card fields into your payment provider's hosted fields also keeps them out of reach of scripts on your own page.
The bigger lesson: every outside script is part of your website
Most business websites load a dozen or more scripts from other companies: analytics, testing, chat, advertising and consent tools. Each one can change at any time without any change on your server. A few habits reduce the risk:
- Keep an inventory of the outside scripts your site loads, and remove the ones you no longer use.
- Keep third-party scripts off login, checkout and account pages wherever you can.
- Use a Content Security Policy with reporting, and watch the reports.
- Limit who can edit tag managers and testing tools, require multi-factor sign-in, and review access regularly.
Indicators of compromise
For your security team: block these addresses, including all subdomains, in DNS filtering, web proxies and firewalls, and search logs from November 13, 2026 onward. Addresses are written with [.] so they cannot be clicked. Compiled from Report URI and StepSecurity.
Domains
1688360[.]cc webshell[.]lol linkforge[.]cc ios-hack[.]icu l7bcsnwc7nzr3uo[.]icu 5vobauua3ojpklx[.]icu p55abnw2pnsz0bo[.]icu eil786npto9u5bq[.]icu 1y69gy2mdzaltz1[.]icu gm38duxoihbmy5o[.]icu q1ls1ho9x7wsxlu[.]icu je4ft3gn8vxudum[.]icu
IP addresses
136.113.106[.]136 (1688360[.]cc) 34.96.213[.]201 (.icu hosts) 34.87.36[.]58 (.icu hosts)
Paths and files
/track.js /html2canvas.min.js /hit.php /card.php /get.js /boot.js /JIUJIU888/index.js /api/iptj
Code markers
window._k7 window.__k7b data-channel-code="02111110" BIGNIAO. The loader was sometimes hidden with atob( (base64), so plain text searches can miss it.
File hashes (SHA-256 of track.js)
814a963c0da480e2449d9b6664480156e1cd6d731014117401bf635badec3a0f f7b84b9a94f749f8ced3f0112d0a66894e8bbc3b4aa5592021d1b8976a9bad40 617df364120adc47fb4f55dcc0c69a9b6c9e108a92c60bc2b64a266254ab02f3. The attacker rebuilt the file 26 times in one day, so hashes alone will miss most versions.
Attacker wallet addresses
| Coin | Address |
|---|---|
| BTC | 19g9rZcPeiz8rUvyWpTRyXXhPnjsWrc3kW, cash-out bc1q5fljq9c4g7gsjgg4etpn6k3lyusahcvjfv0u99 |
| ETH / BSC | 0xC4026b06D82b0F0B363eda0642D61CfbDc17884e |
| TRON | THvs6p56XkUiMkJeSc51cAP64u2RsTYqxa, USDT cash-out TMLQvDvpgSyx94maptc3TxvmCVtT5qHVup |
| SOL | FVtL8NMYS9PA1v3HE67XUGCtGA9JaZUvqav2fU2LrV5n |
| TON | UQAgM1VqSZQiJLSG5w-ovurq5CZdxLEK3_7CMmwALTA5e09p |
| LTC | ltc1q3wp7g6ds0ggg6fl6kncpa4sepn0q3qfacgred5, cash-out ltc1qpxvjs8u8awg8pvpwmssqlj3gzugd9vtwjtc5fc |
| BCH | qzkaaf8n6grmpr0d9lx8xutn3r9lk722f566hhpe58 |
| DOGE | DBEGrxRF7KMSuhTS4c1MFviStU6efdMnu5 |
Other scripts used by the same attacker
Report URI found the same injector (webshell[.]lol/get.js) in scripts from Kissmetrics, StorifyMe and Minute Media between November 13 and 16, and in an SEO "backlink panel" script linked to linkforge[.]cc on Turkish websites since September 17. If your site loads any of these, review them too.
Frequently asked questions
Is it safe to keep using Optimizely?
Optimizely says it has fixed the flaw and secured affected accounts. As of November 17, it had not published how the attacker gained access. Review your projects and access now, add a Content Security Policy, and follow Optimizely's own updates before making a decision.
Our free check says our Optimizely script is clean. Were we safe?
Not necessarily. The check shows what your script contains today. Optimizely removed the injected code on November 16, so a clean result today is expected even for sites that were affected. Your Optimizely change history and your logs are the way to know for sure.
Does this affect other Optimizely products?
Public reports describe Optimizely Web project scripts, the ones loaded into web pages from Optimizely's servers. If you use other Optimizely products, ask Optimizely directly whether they were involved.
Our site does not use Optimizely. Is there anything to do?
Yes. The same kind of attack can come through any outside script. Review which third-party scripts your site loads, keep them off payment and login pages, and add a Content Security Policy.
Want a second pair of eyes?
We can review the outside scripts on your website, build a Content Security Policy that does not break your site, help you meet PCI DSS requirements 6.4.3 and 11.6.1, or investigate whether your site was affected.
Sources
- Report URI (Scott Helme): "World's Largest Brands Infected with Malware via Optimizely," November 16, 2026, updated November 17
- StepSecurity: "OPTIMJACK: Optimizely-Hosted Scripts Loaded a Crypto Clipper and Card Skimmer on 178 Organizations"
- Help Net Security: "Malware spread through Optimizely hits 178 organizations," November 17, 2026
- PCI Security Standards Council: PCI DSS v4.0.1