AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: What To Know About Malware On 178 Major Websites And Its AI Connection on ThorstenMeyerAI.com

Buying for a business?Offer from Amazon

Get business pricing on tech for your team

  • Business-only prices and quantity discounts
  • Tax-exempt purchasing
  • Multiple users, one account, clear invoices
As an affiliate, we earn on qualifying purchases.

TL;DR

Report URI says malicious JavaScript was served to visitors of 178 organizations through Optimizely Experimentation on November 16, 2026. Optimizely says an unauthorized party exploited a vulnerability to inject code into some customer projects; researchers describe attempted theft and exploit-kit delivery, but do not establish that iPhones were compromised.

Malicious JavaScript was served to visitors of websites associated with 178 organizations through Optimizely’s web experimentation platform on November 16, according to security researcher Scott Helme of Report URI. Optimizely confirmed that an unauthorized party exploited a vulnerability to add code to some customer projects, putting visitors at risk of data theft and device-directed attacks without the affected organizations’ own sites necessarily being hacked.

Report URI says it checked 1,611 Optimizely snippets and found malicious code in 188. The researcher’s report counts 178 organizations affected and says the true number may be higher. BleepingComputer separately reported that scans showed the script loaded or attempted to load on sites including Zoom, the BBC, CNN, GitLab, KLM and Salesforce. A listed organization does not mean all its pages or services were affected; some entries were specific subdomains or tools.

The code was placed in the Project JavaScript field in customer projects, according to Report URI. Optimizely’s snippets, which sites load to run page experiments, then delivered the altered code from a source those sites already trusted. Report URI says the rest of each snippet was untouched and that the incident was not a CDN cache-poisoning attack.

Report URI says the script attempted to replace cryptocurrency wallet addresses on pages and in the clipboard, collect payment-card details entered into sites’ own form fields, and capture cookies and screenshots. It also directed some iPhones and iPads toward the Coruna and DarkSword exploit kits. The report says card details entered in hosted payment iframes, such as those provided by Stripe or Adyen, were outside the script’s reach. Hudson Rock cautioned that its iPhone findings show delivery attempts, not confirmed device compromises.

At a glance
reportWhen: Attack observed November 16, 2026; Opti…
The developmentResearchers reported that malicious JavaScript reached visitors to websites associated with 178 organizations through Optimizely Experimentation, and Optimizely confirmed an attacker exploited a vulnerability to alter customer projects.

Optimizely script attack · Nov 13 – Nov 17, 2026

How one trusted script carried malware to 178 websites

The sites were not hacked themselves. The code arrived inside a file they loaded on purpose.

The route (schematic)

Customer projects at Optimizelyvulnerability exploited, per OptimizelyCode written into Project JavaScript
served as usual from Optimizely
The trusted snippetsame address, rest of the file untouchedCarries the attacker’s loader
loaded on every page of
Company websites178 organisations found by Report URI
runs in visitors’ browsers
Visitorswhat the code tried, per Report URI
Swap crypto addressesTake cookies and screenshotsSkim cards in a site’s own formsiPhone exploit kits
All Optimizely files clean by 23:11 UTC

Whether any iPhone was compromised has not been shown.

Timeline (UTC)

  1. Nov 13Earlier wave through three other providers’ scriptsKissmetrics, StorifyMe, Minute Media
  2. Nov 16 04:21First Optimizely injection seen
  3. 08:28Address-swapping payload first seen on a site
  4. 13:43Card skimmer added
  5. 17:13iPhone exploit delivery widened
  6. 19:24Report URI sweep: 186 of 1,611 snippets infected
  7. 21:52Only 1 of 188 infected snippets left
  8. 23:11All infected Optimizely files cleaned
  9. Nov 17 01:02Optimizely confirms the incident on its status page
Why the sites let it in: the code came from the vendor’s own trusted address, so allowing Optimizely did not stop it.

Sources: Report URI (Scott Helme), Nov 16, 2026, timeline and update of Nov 17; Optimizely status page, Nov 17, 2026; Hudson Rock (Infostealers.com), Nov 16–17, 2026. The route is a schematic.

Graphic: Powered by Thorsten Meyer AI

A Trusted Script Reached Many Brands

The incident shows how a compromise at a shared technology provider can reach many unrelated organizations through code they have deliberately installed. In this case, visitors could receive the malicious script from a familiar website even though the point of entry was an Optimizely project, not necessarily the organization’s own infrastructure. That makes the supplier’s access controls and customers’ ability to limit third-party scripts relevant to the security of each affected site.

Report URI also found gaps in defensive policies on the affected organizations’ sites: 24 of 159 blocked the code, while 68 had no policy at all. These figures describe the researcher’s checks of site security policies; they do not establish that every organization without a policy suffered additional damage. Still, the findings indicate that browser-side controls can affect whether injected scripts run, even when a site relies on an external provider.

For visitors, the practical risk depended on the device and activity during exposure. The reported script targeted certain wallet interactions and card-entry fields, and its mobile component attempted to deliver exploit kits. The available reporting does not establish how many people encountered the code, whether any financial information was stolen, or whether any device was successfully compromised.

Products Worth Considering

Bose TV Speaker Compact Soundbar with Bluetooth and HDMI-ARC, Black

Bose TV Speaker Compact Soundbar with Bluetooth and HDMI-ARC, Black

Samsung Odyssey OLED G8 34-Inch Curved Ultra WQHD Gaming Monitor

Samsung Odyssey OLED G8 34-Inch Curved Ultra WQHD Gaming Monitor

As an affiliate, we earn on qualifying purchases.

How the Optimizely Injection Worked

Optimizely Web lets customers test changes to their pages by loading a small JavaScript snippet. Each customer’s snippet can include experiments and optional custom code in a field called Project JavaScript. Report URI says the attacker wrote malicious code into that field across customer projects, allowing the legitimate Optimizely delivery mechanism to serve it to visitors.

Report URI observed the first injection at 04:21 UTC on November 16 and says all infected Optimizely files had been cleaned by 23:11 UTC, a period of about 19 hours. Optimizely’s status-page update, posted at 01:02 UTC on November 17, said it had fixed the vulnerability and removed injected code from affected projects. The company also told BleepingComputer it had ended active sessions, revoked tokens and reset access for affected accounts.

Report URI says the same actor’s injector appeared from November 13 in scripts associated with three other providers—Kissmetrics, StorifyMe and Minute Media—on sites outside the Optimizely list. Hudson Rock also identified that pattern, but the route by which the code entered those services remains unresolved.

“An unauthorized party exploited a vulnerability in Experimentation and used it to add malicious JavaScript to some customer projects.”

— Optimizely, in its status-page update

The Access Path and Damage Remain Unknown

Optimizely has said the vulnerability was in the Experimentation account signup flow, but its public statements do not explain how that flaw enabled changes across many unrelated customer projects. Report URI says it does not know how the attacker obtained write access, and Hudson Rock says the company’s disclosure does not explain the bug’s code-level mechanics.

It is also not clear how many visitors received the code or whether it resulted in confirmed theft of credentials, payment details, cryptocurrency or other data. Hudson Rock’s reporting does not establish successful iPhone compromises. The list of 178 organizations indicates identified affected organizations, not that each organization’s entire website, checkout system or customer base was exposed. Report URI says its scan count suggests the total number may be higher, but the comparison basis for that estimate is not specified.

Investigations and Customer Checks Continue

Optimizely says it has fixed the vulnerability, removed injected code from affected projects, notified affected customers and reset access for affected accounts. The next public information that could clarify the incident would include a fuller technical account of the signup-flow vulnerability, how access to customer projects was obtained and whether the company found evidence of data theft or successful device compromise.

Organizations named in the research will need to determine which specific projects and pages loaded the code and assess whether visitors entered information during the exposure window. Visitors concerned about possible iPhone targeting should check their software version: Report URI advises updating to iOS 18.7.3 or later, or iOS 26.3 or later. The report says both versions had been released months earlier by Apple. The source material does not say whether further updates to the affected-site list or investigation findings have been announced.

Key Questions

What happened on November 16?

Report URI says malicious JavaScript was served through Optimizely Experimentation to visitors of sites associated with 178 organizations. Optimizely confirmed an unauthorized party exploited a vulnerability to add code to some customer projects.

Were all pages of the named organizations affected?

No. Affected entries may be specific subdomains or tools. Hudson Rock cautioned that an affected subdomain does not establish exposure across an organization’s entire website, products, customer base or checkout system.

Was the malware linked to artificial intelligence?

The source material does not report an AI connection. It describes malicious JavaScript delivered through a web experimentation platform and does not attribute the code, attack or vulnerability to artificial intelligence.

Were iPhones confirmed to be compromised?

No successful iPhone compromise is confirmed in the cited findings. The code reportedly directed some iPhones and iPads toward exploit kits, but Hudson Rock describes its evidence as delivery observations, not demonstrated device compromises.

What should visitors do?

Report URI advises iPhone users to update to iOS 18.7.3 or later, or iOS 26.3 or later. The source does not provide a confirmed count of affected visitors or confirmed cases of stolen data.

Source: ThorstenMeyerAI.com

You May Also Like

Anthropic Gives Vetted Defenders Fewer Claude Guardrails – Dark Reading

Dark Reading reports fewer Claude guardrails for vetted defenders, but eligibility, affected safeguards and rollout details are not available.

Dessa: The Most Authentic Full-duplex Speech Model Ever Built

Block’s Dessa page describes a full-duplex speech model, but provides no technical results, release details or evidence for its authenticity claim.

Learning Or Building? 12 AI Coding Assistants To Consider In 2027

A 12-book guide compares broad AI programming resources with beginner, governance, tool-specific and project-focused titles.

The Agent Economy’s Plumbing: Four AI Deals In One Day

Four October 29 deals and launches target voice infrastructure, agent identity, building design and personalized shopping. Several details remain undisclosed.