Optimizely Compromised, Served Malware On Zoom, GitLab
AIThis post was created with the assistance of artificial intelligence (AI).

TL;DR

Report URI said malicious JavaScript was served through Optimizely Web snippets used by 178 organisations on Nov. 16, including Zoom, GitLab, the BBC and CNN. The payload could steal cryptocurrency and payment-card data, capture cookies and screenshots, and deliver iOS exploit kits. The attacker’s route into customer projects and the number of affected visitors remain unknown.

Malicious JavaScript was served to visitors of 178 organisations through the Optimizely Web experimentation platform on Nov. 16, according to cybersecurity company Report URI. The affected sites included Zoom, GitLab, the BBC, CNN, Salesforce, IKEA and HSBC; the injected code could steal cryptocurrency and payment-card information, capture cookies and screenshots, and deliver iOS exploit kits.

Before you orderOffer from Amazon

Get privacy and security gear delivered free with Prime

  • Fast, free delivery on millions of items
  • Prime Video, Amazon Music and more included
  • Member-only deals all year
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

Report URI said attackers altered the Project JavaScript field in customer projects. Optimizely Web snippets load from Optimizely’s CDN, and the field allows customers to add custom code that runs on their pages. The attackers’ additions caused browsers to load a separate script from 1688360.cc. The affected snippets continued to be served from trusted Optimizely domains, according to the report.

The report described three forms of the injected code: a loader pointing to the external script, a version that concealed the address using base64, and an inline copy of the payload that could run if the external domain was blocked. Report URI said some organisations hosted Optimizely snippets on their own domains, but those snippets were also affected because their content came from the same Optimizely projects.

The payload included a cryptocurrency address replacer and a payment-card skimmer, and could steal cookies and take screenshots. On iPhones and iPads, Report URI said it delivered the Coruna and DarkSword iOS exploit kits. It also reported verified cryptocurrency transactions to the attackers. The company said the code changed repeatedly during the day as features and wallet addresses were added.

In its final reported sweep, at 21:52 UTC, Report URI found the malicious code in one of 188 snippets it checked, down from 169 in that set at 20:56 UTC. The remaining snippet was associated with TalkTalk. These figures describe the snippets examined in those checks, not the number of affected visitors or the full population of Optimizely customers.

At a glance
reportWhen: Reported Nov. 16, 2026; most observed s…
The developmentReport URI reported that an attacker injected malicious JavaScript into Optimizely Web customer projects, causing affected sites to load malware from a third-party domain.

How a Trusted Script Reached Visitors

The incident shows how a compromise of a third-party web service can expose visitors even when an organisation’s own website code and infrastructure have not been breached. Browsers on affected sites were directed to run attacker-controlled code through a snippet the sites already trusted. That makes the incident relevant to both the organisations using the platform and people who visited their pages during the exposure window.

The reported payload covered several kinds of harm: financial theft through substituted wallet addresses and card skimming, possible account exposure through stolen cookies, and device compromise attempts through iOS exploit kits. Report URI said cryptocurrency had already been stolen, citing verified transactions. It did not provide a total value, a count of victims, or a confirmed tally of stolen card or account data.

The event also highlights the limits of relying on a script’s hosting domain as a sign of safety. The affected code was delivered through Optimizely infrastructure trusted by the sites, but the report says the malicious changes were made in customer project settings. Monitoring for unexpected script changes and restricting browser permissions through an enforced content security policy may help detect or limit similar activity, but Report URI’s discussion of those controls is guidance from the security company, not evidence that every affected organisation had them in place.

Products Worth Considering

LG UltraGear 27-inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDR True Black 400

LG UltraGear 27-inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDR True Black 400

Apple AirPods Pro 3 Wireless Earbuds with Active Noise Cancellation

Apple AirPods Pro 3 Wireless Earbuds with Active Noise Cancellation

As an affiliate, we earn on qualifying purchases.

How Optimizely Snippets Were Changed

Optimizely Web lets customers run experiments by placing a JavaScript snippet on their websites. According to Report URI, each customer’s snippet can include Project JavaScript, a field for custom code entered through the Optimizely dashboard. The attack modified that field, leaving the rest of the snippet intact and continuing to serve it from the expected host.

Report URI’s timeline says it first observed an Optimizely injection on the BBC’s snippet at 04:21 UTC on Nov. 16; an earlier clean capture was recorded at 02:21 UTC. It later saw a card skimmer added to the external script and iOS exploit-kit delivery introduced during the afternoon. The report also identified an earlier wave involving the same injector through other third-party scripts on sites including securitybank.com, faz.net and smh.com.au. That history does not establish that those earlier incidents shared the same access route into Optimizely projects.

Report URI explicitly said this was not a CDN cache-poisoning attack: in its account, Optimizely’s CDN served the content it had been given. The company also reported that Optimizely’s own websites had snippets carrying the code. These findings are based on Report URI’s observations; the source material does not include a public technical account from Optimizely explaining how the project changes occurred.

“The important lesson here is that trusted JavaScript doesn’t always remain trustworthy.”

— Report URI

Access Route and Victim Impact

The route used to change customer projects remains unknown. Report URI said it could not determine whether attackers compromised an Optimizely system, obtained credentials that allowed access across customers, or used another method. The report leaves that explanation for Optimizely to provide.

The full scope of harm is also unclear. Report URI described verified cryptocurrency theft but supplied no total value, and the source material does not establish how many visitors encountered the payload, how many devices were affected, or whether payment-card details, cookies or screenshots were successfully taken. The report’s snippet counts refer to particular scans and should not be read as visitor or victim counts.

Report URI said most malicious code had been removed by the time of its late-evening checks, but its observation of one remaining infected snippet means that cleanup was not complete in the evidence it presented. The source material does not confirm whether every affected project or organisation had been checked and remediated.

Optimizely’s Investigation and Cleanup

The next key development is an account from Optimizely about how attackers gained write access to the customer projects and whether the issue extended beyond the snippets Report URI inspected. The source material does not include an Optimizely statement, a published incident timeline from the company, or a confirmed explanation of its investigation.

Organisations that use Optimizely Web will need to establish whether their projects or snippets were affected, remove unauthorised code, and review activity during the exposure period. Visitors concerned about financial or account exposure can monitor relevant payment and cryptocurrency activity, though the report does not identify individual victims or provide a complete list of confirmed losses. Further information is needed to establish the final scope, the duration of exposure for each site, and whether stolen data was used beyond the cryptocurrency transactions Report URI described.

Key Questions

Which organisations were affected?

Report URI said the injected code reached visitors to 178 organisations, including Zoom, GitLab, the BBC, CNN, Salesforce, IKEA, HSBC, Equifax, Air France-KLM, Ledger and Peacock. The report does not provide a confirmed count of affected visitors.

How did the malicious code reach the websites?

According to Report URI, attackers added code to the Project JavaScript field in Optimizely Web customer projects. The resulting snippets continued to load through Optimizely or customer-hosted domains, then directed browsers to an external malware script.

What could the malware do?

The payload could replace cryptocurrency wallet addresses, skim payment-card information, steal cookies and capture screenshots. Report URI also said it delivered Coruna and DarkSword iOS exploit kits to some visitors using iPhones and iPads. The report does not confirm how many people experienced each capability.

Has the threat been removed?

Report URI said the code had been removed from most snippets it checked. At 21:52 UTC on Nov. 16, it found one infected snippet among 188 in its sweep. That observation does not establish that all affected projects had been checked or cleaned.

Do investigators know how the attackers gained access?

No. Report URI said it could not determine whether the attackers compromised an Optimizely system, used credentials with access across customer projects, or obtained access another way. The source material does not include Optimizely’s explanation.

Source: blog.report-uri.com

You May Also Like

OpenAI Engineer Arrested In Bugging Of Executive Conference Room

Federal prosecutors say an OpenAI engineer hid a recording device in an executive conference room. He is charged and remains in custody.

One Hacker, A Swarm Of AI Agents, 100+ Company Websites: What Changes When A Break-In Costs About $50

Reuters and NorForge describe an operator using Cairn AI agents to probe more than 1,200 sites and breach websites tied to over 100 companies.

Is Your Chrome Version Safe From The XML Bug Attackers Are Using?

Google says attackers are exploiting a high-severity Chrome XML bug. Check your browser version and update when a fix is available.

The Sony Pictures Hack: When a Comedy Movie Triggered a Cyberwar

Uncover how a comedy film led to a major cyberattack, revealing unexpected motives behind the Sony Pictures hacking incident.