SpeedyGoSpeedyGo: One plugin to make your WordPress website faster
Try Now
CorvexaCorvexa: AI that turn more website visitors into qualified leads.
Try it instantly
Menu
logologo+1-256-548-8850
TopDesignKing
back_iconRead More
Back to blog page

Biggest Cyberattack of 2026: Hackers Hit 178 Major Organizations Including BBC, CNN and Zoom

News & Articlesdate_icon 19/11/2026
Biggest Cyberattack of 2026: Hackers Hit 178 Major Organizations Including BBC, CNN and Zoom
Quick Summary

A cyberattack involving Optimizely, a popular website experimentation platform, caused malicious JavaScript to appear on websites associated with 178 organizations, including BBC, Zoom, CNN, HSBC, Salesforce, and GitLab.

The attackers didn’t need to break into each company’s website. Instead, they exploited a vulnerability in Optimizely and used its trusted JavaScript delivery system to distribute harmful code.

The malicious script was designed to replace cryptocurrency wallet addresses, steal payment card information, collect browser data, and attempt to exploit older iPhones and iPads.

Security researchers discovered 188 affected JavaScript files among 1,611 files examined. Although the infected scripts were eventually cleaned, the incident raised serious questions about how much control third-party website tools have over visitors’ browsers.

What Happened in the Optimizely Cyberattack?

Imagine opening the BBC website on a Monday morning.

The website looks completely normal. The address is correct, the connection is secure, and everything appears to be working.

But something is happening in the background. Your browser is running JavaScript that was added by an attacker.

BBC website screenshot showing malicious JavaScript code running on a trusted website with a secure HTTPS padlock.

A trusted website can still run harmful JavaScript if attackers compromise a third-party script.

The same problem appeared on websites and services associated with major organizations, including:

  • BBC, CNN, Fox News, and NPR
  • Zoom and Salesforce
  • HSBC and Equifax
  • GitLab and Smartsheet
  • IKEA and KLM
  • Ledger, a cryptocurrency hardware wallet company
Logos of BBC, Zoom, CNN, Fox News, NPR, HSBC, Equifax, Salesforce, GitLab, IKEA, KLM, and Ledger shown as organizations affected by the Optimizely JavaScript supply-chain attack.

Major websites, including BBC, CNN, Zoom, and Salesforce, were exposed to malicious code through a compromised third-party service.

At first, it might sound like hackers broke into all these companies at once. But that isn’t what happened.

Most of the affected organizations had not suffered a confirmed breach of their internal systems.  Instead, the attackers targeted a third-party service that these websites trusted. That service was Optimizely.

When visitors opened affected pages, their browsers could automatically load malicious code through Optimizely’s JavaScript file. This is known as a software supply chain attack.

Rather than attacking every website separately, the attacker compromised a shared service used by many websites.

Did the attack really affect 178 organizations?

Infographic showing 178 organizations affected by the Optimizely JavaScript supply-chain attack, based on research by Scott Helme of Report URI.

The figure came from security researcher Scott Helme, who runs Report URI.

Screenshot of Scott Helme's Report URI article titled "World's Largest Brands Infected With Malware Via Optimizely," published November 16, 2026.

Security researcher Scott Helme published his findings on Report URI, revealing how major brands were exposed to malware through Optimizely.

His team examined 1,611 copies of Optimizely JavaScript files. Of those, 188 contained malicious code, associated with 178 organizations.

The researchers believed the actual number could be higher because they could only examine the files they managed to collect. However, being on the affected list didn’t mean an organization’s entire website was compromised.

For example, the affected IKEA entry involved a particular home design tool rather than every IKEA page.

Screenshot showing IKEA's client.home-design.ikea.com domain listed among websites affected by the November 2026 Optimizely JavaScript supply-chain attack.

The Optimizely attack affected IKEA’s home design tool domain, not necessarily the company’s entire website.

Equifax also said it had found no evidence that its own systems had been compromised. Reuters reported that the company stopped using Optimizely as a precaution.

Illustration showing harmful JavaScript running in a visitor's browser while the company's internal network remains secure during the Optimizely cyberattack.

Malicious code could run in visitors’ browsers without attackers breaking into the affected company’s internal network.

So the distinction matters.

The attack reached websites belonging to 178 organizations through malicious third-party code. It did not establish that hackers had gained access to 178 corporate networks.

Infographic showing how one malicious Optimizely JavaScript delivery reached websites associated with 178 organizations.

A single compromised third-party service distributed malicious code across websites belonging to 178 organizations.

Infographic displaying Report URI's Optimizely cyberattack findings: 1,611 JavaScript files checked, 188 infected, and 178 organizations affected.

Report URI researcher Scott Helme examined 1,611 Optimizely JavaScript files and found malicious code in 188 files associated with 178 organizations.

What Is Optimizely, and Why Do So Many Websites Use It?

Optimizely A/B testing tool illustration showing two website versions displayed to 50% of visitors each to identify which design performs better.

Optimizely helps businesses compare two versions of a webpage to see which one performs better and generates more conversions.

To understand how this happened, we first need to understand what Optimizely does.

Have you ever visited an online store and noticed that a button, headline, or page layout looks different from your previous visit?

That could be an A/B test.

For example, a business might want to know which button gets more clicks.

One group of visitors sees a green button. Another group sees an orange button. The company measures the results and chooses the version that performs better.

Optimizely helps businesses run these kinds of experiments without changing the website’s main code every time.

According to Optimizely’s own website, more than 2 million experiments have been run using its platform.

How does Optimizely work on a website?

A website adds a small JavaScript reference to its pages. That reference tells the visitor’s browser to download a file from Optimizely’s servers. The file is commonly called the Optimizely snippet.

Each customer has its own snippet, which contains the instructions needed to run that customer’s experiments. Those instructions might tell the browser to change a button color, replace a headline, adjust a layout, or show different content to different visitors. The useful part for marketers is that they can make these changes through the Optimizely dashboard. They don’t need a developer to update the website’s source code for every experiment. But this convenience also creates a security risk.

The Optimizely snippet has permission to make changes directly inside the visitor’s browser. And because it needs to modify a page before visitors notice the original version, Optimizely recommends loading the snippet very early in the HTML document. Usually, it is placed near the top of the <head> section. Otherwise, visitors might briefly see the original page before the experiment changes it.

This effect is called flicker.

So the Optimizely script runs early, has access to the webpage, and can be updated from a separate dashboard.

That combination became important during the attack.

How Did the Attackers Insert Malicious Code Into Optimizely?

Optimizely has an advanced feature called Project JavaScript.

It allows users to add custom JavaScript to their Optimizely projects. mThe code entered there can run before the normal web experimentation code. For legitimate users, this helps with advanced experiments and custom website behavior. But attackers found a way to abuse this feature.

According to Report URI, malicious JavaScript was inserted into the Project JavaScript configuration of affected Optimizely projects.

The added code instructed visitors’ browsers to load another JavaScript file from a server controlled by the attacker. The original Optimizely snippet was otherwise left largely unchanged. That made the attack harder to notice. The file was still delivered from Optimizely’s legitimate servers using a valid HTTPS connection. To the browser, it looked like an ordinary script requested by the website. The malicious instructions had been added at the source

Optimizely Cyberattack Timeline: What Happened Throughout the Day?

The attack developed over several hours.

The following timeline uses UK time, matching the security researchers’ observations described in the report according to URI.

2:21 AM: BBC’s Optimizely file was still clean
An internet scanning service captured a copy of the BBC’s Optimizely JavaScript file. At this point, researchers found no malicious code in that copy.

Timeline of the November 2026 Optimizely cyberattack showing the BBC's last clean JavaScript capture at 02:21 UTC and the first detected malicious injection at 04:21 UTC.

The BBC’s Optimizely JavaScript was clean at 02:21 UTC, but researchers detected malicious code by 04:21 UTC on November 16, 2026.

4:21 AM: Malicious code appeared
A later copy contained additional JavaScript. The new instructions told visitors’ browsers to fetch code from an attacker-controlled website.

Optimizely cyberattack timeline showing the BBC's JavaScript file containing malicious code at 4:21 AM UK time on November 16, 2026.

By 4:21 AM UK time on November 16, 2026, the BBC’s Optimizely JavaScript file was no longer clean and contained attacker-injected code.

5:23 AM: A new attacker domain was registered
The attacker registered another web address, which became the main location for distributing the malicious script

Optimizely cyberattack timeline showing a new attacker-controlled domain registered at 5:23 AM on November 16, 2026, to distribute malicious JavaScript.

At 5:23 AM, the attackers registered a new domain that became the main location for distributing malicious JavaScript throughout the day.

8:10 AM: Main malicious script appeared
Researchers observed the first version of the main payload. Its initial purpose was to interfere with cryptocurrency transactions.

Optimizely cyberattack timeline showing the first version of the main malicious JavaScript payload at 8:10 AM UK time, initially targeting cryptocurrency transactions.

At 8:10 AM on November 16, 2026, researchers observed the first version of the main malicious script, which initially focused on cryptocurrency theft.

8:28 AM: BBC was serving the malicious code

The compromised JavaScript was observed running through bbc.com.

Optimizely cyberattack timeline showing BBC.com serving attacker-injected malicious JavaScript at 8:28 AM UK time on November 16, 2026.

By 8:28 AM on November 16, 2026, BBC.com was observed serving the malicious JavaScript through its compromised Optimizely script.

11:20 AM: Zoom’s Optimizely file was modified
Other major organizations, including Salesforce, CNN, Fox News, GitLab, and Smartsheet, were also associated with affected scripts.

Optimizely cyberattack timeline showing Zoom's JavaScript file modified at 11:20 AM, alongside affected organizations including BBC, CNN, Salesforce, Fox News, GitLab, and Smartsheet.

At 11:20 AM on November 16, 2026, Zoom’s Optimizely file was observed to be modified, while several other major organizations were also associated with affected scripts.

1:43 PM: Payment card skimming was added
The attacker updated the script to look for payment card details entered into accessible webpage forms.

Optimizely cyberattack timeline showing payment card skimming functionality added to the malicious JavaScript at 1:43 PM, alongside an earlier cryptocurrency theft feature.

At 1:43 PM on November 16, 2026, attackers updated the malicious script to target payment card information entered into accessible webpage forms.

5:00–6:00 PM: iPhone attack functionality appeared
Another component was added to target certain iPhone and iPad software versions.

Optimizely cyberattack timeline showing malicious JavaScript updated between 5 PM and 6 PM to target certain iPhone and iPad software versions.

Between 5 PM and 6 PM on November 16, 2026, attackers added a component targeting certain iPhone and iPad software versions.

7:24 PM: Report URI completed a major scan}
Researchers examined 1,611 Optimizely files and found 186 infected at that point.

Optimizely cyberattack timeline highlighting Report URI's major scan at 7:24 PM on November 16, 2026.

At 7:24 PM on November 16, 2026, Report URI completed a major scan to investigate the spread of malicious Optimizely JavaScript.

8:10 PM: Scott Helme publicly warned about the incident
He posted on X about a third-party JavaScript compromise affecting hundreds of high-value websites.

Screenshot of Scott Helme's X post on November 16, 2026, at 8:10 PM UK time, warning about a third-party JavaScript breach affecting hundreds of major websites

At 8:10 PM UK time, security researcher Scott Helme publicly warned about a third-party JavaScript compromise affecting hundreds of high-value websites.

8:40–9:07 PM: Major cleanup began
Ledger’s affected file was cleaned around 8:40 PM, Smartsheet’s around 8:56 PM, and the BBC’s around 9:07 PM.

Optimizely cyberattack cleanup timeline showing affected websites being cleared of malicious JavaScript between 8:40 PM and 9:07 PM on November 16, 2026.

Between 8:40 PM and 9:07 PM, researchers observed malicious JavaScript being removed from affected website files as the Optimizely attack cleanup began.

9:32 PM: Only 41 affected files remained
Report URI said the number of infected files had fallen significantly.

Screenshot of Scott Helme's November 16, 2026 X post confirming that only 41 of 188 infected Optimizely JavaScript files remained compromised at 9:32 PM UK time.

At 9:32 PM, security researcher Scott Helme reported that only 41 of the 188 infected Optimizely JavaScript files remained affected as cleanup progressed.

9:52 PM: One infected file remained
The remaining affected file was associated with TalkTalk.

Optimizely cyberattack timeline showing TalkTalk as the last affected company at 9:52 PM UK time, while other website files had been cleaned.

By 9:52 PM on November 16, 2026, TalkTalk was the only organization shown with an infected file remaining in the cleanup timeline.

11:11 PM: All identified files were clean
Report URI reported that the identified infected Optimizely files had been cleaned.

Optimizely cyberattack timeline showing all identified infected JavaScript files cleaned by 11:11 PM UK time on November 16, 2026.

By 11:11 PM on November 16, 2026, researchers reported that all identified infected Optimizely JavaScript files had been cleaned.

The attacker kept updating the script

This wasn’t a case where the attacker uploaded one malicious file and left it alone. Report URI identified 26 builds of the main script during the day.

Infographic showing 26 versions of the main malicious JavaScript created during the November 2026 Optimizely supply-chain cyberattack.

Report URI identified 26 builds of the main malicious script as attackers updated and expanded its capabilities throughout the day.

Each new version added functionality or expanded the attack. The attacker introduced additional cryptocurrency wallets, payment card theft, and iOS exploitation capabilities. Step Security estimated that most affected sites served malicious code for approximately 16 hours.

Infographic showing the November 2026 Optimizely JavaScript supply-chain cyberattack lasted approximately 16 hours, with a timeline comparing the attack duration to a full 24-hour day.

The Optimizely supply-chain attack lasted around 16 hours, during which malicious JavaScript affected websites belonging to major organizations.

Some websites were infected again after cleanup

Smartsheet was one example. Its affected file was cleaned, but researchers later observed malicious code appearing again.

Before-and-after illustration showing Smartsheet's Optimizely JavaScript file cleaned of malicious code and then infected again during the November 2026 cyberattack.

Smartsheet’s affected JavaScript file was initially cleaned, but malicious code appeared again, showing that the Optimizely attack was not fully resolved.

This time, the attacker had disguised the address of the external script by encoding it. The change suggested that the attacker still had access to the configuration after the initial cleanup.

There was another complication. On some websites, including BBC and one HSBC file, the attacker embedded a complete copy of the malicious payload directly inside the Optimizely snippet. That backup contained roughly 24 KB of code. If the external malicious script failed to load within approximately 2.5 seconds, the embedded version could run instead.

This meant blocking the attacker’s external domain alone was not enough to stop every part of the attack.

What Did the Malicious JavaScript Actually Do?

The attack was especially serious because it combined several different techniques.

According to Report URI, the script contained four major capabilities, with multiple components active in the same payload.

1. It Replaced Cryptocurrency Wallet Addresses

The first capability targeted cryptocurrency transactions. When someone sends cryptocurrency, they usually copy a wallet address and paste it into a payment form. These addresses are long combinations of letters and numbers. Most people don’t type them manually.

Illustration of a long cryptocurrency wallet address made up of letters and numbers, showing how attackers exploited copied wallet addresses during the Optimizely cyberattack.

Cryptocurrency wallet addresses contain long strings of letters and numbers. Attackers exploited this by using malicious JavaScript to replace legitimate wallet addresses with their own.

The malicious script took advantage of that behavior. It searched webpages for cryptocurrency wallet addresses and replaced them with addresses controlled by the attacker. The script could interfere with addresses appearing in page text, form fields, links, clipboard operations, and outgoing browser requests.

For example, someone might copy the correct Bitcoin address from a website.

But when they pasted it, the address could have been replaced with the attacker’s wallet. If the person confirmed the transaction without noticing the change, the money could go to the attacker instead.

Why wouldn’t the wallet detect the wrong address?

Cryptocurrency addresses often include mechanisms that help detect typing mistakes. But the attacker wasn’t inserting random characters. The script replaced the original address with another valid wallet address. So the address could pass normal validation checks even though it belonged to the wrong person.

That is why checking only whether an address looks valid isn’t enough.

2. It Tried to Steal Payment Card Information

Later in the day, the attacker added payment card skimming functionality. This part of the script looked for sensitive card information entered into webpages. That included card numbers, expiry dates, and card security codes. When the script found accessible payment details, it could send them to an attacker-controlled server. This kind of attack is commonly called web skimming or Magecart-style card skimming. However, the attack couldn’t read every payment form.

Many websites use payment providers such as Stripe or Adyen. These providers can place sensitive payment fields inside isolated frames hosted on their own domains. Because browsers enforce restrictions between different website origins, JavaScript running on the main shopping page generally cannot read inside those protected frames. That created an important limitation.

Cards entered into properly isolated payment fields were outside the reach of this particular skimmer. Cards entered directly into accessible fields on the affected website could be exposed.

3. It Collected Screenshots, Cookies, and Browser Information

The malicious script also collected information about what visitors were doing. When it detected certain wallet or payment-related activity, it could send additional information to the attacker. That included the webpage address, readable cookies, and a screenshot of the page.

Illustration showing malicious JavaScript collecting a webpage address, browser cookies, and a screenshot during the Optimizely cyberattack.

The malicious Optimizely script could send webpage addresses, certain browser cookies, and screenshots back to attacker-controlled servers.

A screenshot could reveal information visible on the screen, while cookies and browser data could provide additional context about a visitor’s session.

Could the attacker steal login sessions?

Potentially, some browser cookies could be exposed. But not all cookies are accessible to JavaScript. Websites can mark sensitive cookies with the HttpOnly attribute. That prevents ordinary webpage JavaScript from reading those cookies. Report URI said cookies protected in this way were not accessible to the malicious script.

This is one reason secure cookie settings are important.

4. It Attempted to Exploit Older iPhones and iPads

The fourth capability targeted Apple devices. When someone visited an affected website using an iPhone or iPad, the malicious script checked the device’s iOS version. It then attempted to load exploit code through a hidden frame.

Illustration showing how malicious JavaScript checked an iPhone's iOS version, loaded a hidden frame, and attempted to exploit security vulnerabilities during the Optimizely cyberattack.

The malicious script checked the visitor’s iOS version and attempted to load an exploit through a hidden frame to target vulnerable iPhones.

The attack used two exploit kits identified as Coruna and DarkSword.

Infographic displaying Coruna and DarkSword, two iPhone exploit kits associated with the November 2026 Optimizely JavaScript supply-chain attack.

The attackers incorporated components associated with Coruna and DarkSword, two exploit kits designed to target vulnerabilities in certain iPhone software versions.

These were not ordinary tracking scripts. They were collections of exploits designed to take advantage of security vulnerabilities in certain versions of iOS.

What is the Coruna exploit kit?

Infographic showing Google Threat Intelligence Group's March 2026 research into the Karuna iPhone exploit kit and mobile security vulnerabilities.

Google Threat Intelligence Group investigated the Karuna iPhone exploit kit in March 2026, before it was linked to the November Optimizely cyberattack.

Google’s Threat Intelligence team had previously investigated Coruna.

According to the research according to Report URI, Coruna targeted iOS versions 13 through 17.2.1. The kit contained 23 exploits.

Google had observed it moving between different types of operators, including surveillance-related users, espionage actors, and financially motivated criminals. The research showed how sophisticated exploitation tools can move beyond their original operators and eventually become available to other attackers.

What is the DarkSword exploit kit?

DarkSword was another iOS exploitation toolkit. It targeted iOS versions 18.4 through 18.7.

Infographic comparing the iOS versions targeted by the Coruna and DarkSword iPhone exploit kits, including iOS 13–17.2.1 and iOS 18.4–18.7.

Coruna targeted iOS versions 13–17.2.1, while DarkSword targeted iOS 18.4–18.7, showing the different software versions associated with these exploit kits.

Google, Lookout, and iVerify published research about the kit in March 2026. Their findings linked its use to multiple surveillance-related operators and suspected state-sponsored activity.

TechCrunch subsequently reported that a newer version of the toolkit had leaked publicly. That made the situation more concerning.

Tools previously associated with targeted attacks were now being delivered through JavaScript on widely visited websites.

Were iPhones actually hacked during this incident?

This is where the evidence needs to be separated from the attacker’s intentions.

Researchers confirmed that the malicious code contained mechanisms to deliver iOS exploit kits. But there was no public confirmation in the source material that a real user’s iPhone had been successfully compromised through this campaign.

Some researchers tested the delivery using tools that simulated iPhone devices.

Hudson Rock emphasized that demonstrating delivery of exploit code is not the same as proving a successful device compromise.

The malicious script also behaved differently depending on the iOS version.

On devices running iOS 18.6 or older, researchers observed code that could reload the page every 60 seconds to try again.

Infographic showing how malicious JavaScript could reload a webpage every 60 seconds on iPhones running iOS 18.6 or older during the Optimizely cyberattack.

Researchers observed code designed to reload webpages every 60 seconds on devices running iOS 18.6 or older, potentially allowing repeated exploitation attempts.

For newer iPhones, the script appeared to collect device information rather than launch the same exploitation sequence. The source identified iOS 18.7.3 and iOS 26.3 as versions containing relevant fixes for the targeted vulnerabilities.

Infographic comparing Coruna and DarkSword iPhone exploit kits, showing targeted iOS versions and security patches in iOS 18.7.3 and iOS 26.3.

Coruna and DarkSword targeted vulnerabilities in older iOS versions. Apple addressed relevant security issues in iOS 18.7.3 and iOS 26.3.

The practical advice is to keep iPhones and iPads updated to the latest compatible security release.

How Much Money Did the Attackers Steal?

Infographic showing approximately $7,040 received by cryptocurrency wallets linked to the Optimizely cyberattack, including Bitcoin, Litecoin, Tron, Ethereum, and BNB Chain.

By 19:25 UTC on November 16, 2026, wallets associated with the attack had received approximately $7,040 across five cryptocurrencies, although not all funds were confirmed stolen from victims.

The cryptocurrency component offered one way for researchers to estimate the attack’s financial impact. Blockchain transactions are publicly visible, so investigators could examine money received by wallets associated with the malicious script. However, different researchers arrived at different figures.

Source Reported amount
Report URI Approximately $7,040
Step Security Approximately $6,600
Cointelegraph blockchain review Approximately $7,100
Galaxy Digital’s Alex Thorn, as reported by Cointelegraph Approximately $2,300 potentially attributable to victims

The difference matters because money received by an attacker-controlled wallet isn’t automatically stolen money. Some transactions may involve funds being moved between wallets controlled by the same person. Alex Thorn suggested that a significant portion of the observed Bitcoin activity could have been the attacker moving their own funds. That means the amount actually stolen through cryptocurrency address replacement may have been considerably lower than the total received.

What happened to the money?

Early the following Tuesday, between approximately 4:00 AM and 5:30 AM UK time, the attacker-controlled wallets were emptied.

Report URI observed that one of the outgoing transfers went to an address previously associated with other suspicious activity. The address had appeared in operations during March 2026 and on November 3, 2026. This suggested a possible connection between the Optimizely campaign and earlier attacks.

However, the researchers did not publicly identify the other operations involved.

What about stolen cards and personal information?

Those losses were much harder to measure. Unlike cryptocurrency transactions, stolen payment card details, screenshots, and cookies do not appear on a public blockchain.

Report URI said it could not determine how many cards had been captured, how many screenshots or cookies were stolen, or how many iOS devices might have been compromised. So the publicly observable cryptocurrency totals should not be treated as the full impact of the incident.

How Did the Attackers Gain Access to Optimizely?

This was one of the biggest unanswered questions.

How could an attacker modify the Project JavaScript configuration of so many customer projects?

Optimizely later acknowledged that an unauthorized party had exploited a vulnerability in its experimentation platform.

According to its statement, the attacker used the vulnerability to insert malicious JavaScript into some customer projects.

Optimizely said the weakness was related to the account sign-up flow for its experimentation product.

That is normally the process used when someone creates a new account. But the exact connection between the sign-up vulnerability and the ability to modify other customers’ projects was not fully explained in the source’s reporting.

What did Optimizely do after discovering the attack?

Optimizely said it had taken several steps to contain the incident. These included fixing the vulnerability, removing malicious JavaScript, securing affected accounts, terminating active sessions, revoking access tokens, resetting access, and notifying affected customers. The company also continued investigating the incident.

At the time covered by the video, its status page was still in a monitoring phase. Researchers from Report URI, Step Security, and Hudson Rock noted that the full technical explanation of the initial access method remained incomplete.

The Attack May Have Started Before the Optimizely Incident

The Optimizely campaign was the largest wave according to Report URI, but researchers found related activity involving other services.

Starting on November 13, Hudson Rock observed malicious JavaScript appearing through three additional third-party platforms.

These were Kissmetrics, a website analytics service; Storify Me, a story-style content widget; and Minute Media, a provider of media-related scripts.

Through these services, related malicious code appeared on websites associated with organizations including a bank in the Philippines, Frankfurter Allgemeine in Germany, the Sydney Morning Herald, VidaXL, and Fox Sports.

Infographic showing how compromised third-party scripts from Kissmetrics, StorifyMe, and Minute Media affected websites including Security Bank, FAZ, Sydney Morning Herald, Vidaxl, and Fox Sports.

Researchers identified malicious changes involving Kissmetrics, StorifyMe, and Minute Media scripts, exposing websites that relied on these third-party services.

Were all these attacks carried out by the same person?

Researchers found similarities between the campaigns.

A researcher operating the website Ohe identified a shared tracking identifier, described as a channel number, across the different waves. That suggested the same exploitation toolkit or service may have been involved.

Ohe believed one buyer of the toolkit could have operated all the waves.

Hudson Rock was more cautious. It said shared code could establish a relationship between the campaigns without proving that the same individual controlled them or used the same initial access method.

The methods used to compromise the other three services were also not established in the source material.

Who Was Behind the Cyberattack?

At the time covered by the video, the attacker’s identity remained unknown.

Researchers had identified some technical clues.n One researcher reported that the iPhone exploit kit’s control panel contained Chinese-language elements. Some infrastructure was associated with servers in Hong Kong and Singapore.

But those details were not proof of the attacker’s nationality or identity.

Infographic showing possible clues about the Optimizely cyberattack, including a Chinese-language exploit kit control panel and servers in Hong Kong and Singapore, without confirming the attacker's identity.

Researchers found Chinese-language elements and infrastructure linked to Hong Kong and Singapore, but these clues did not prove who was responsible for the attack.

Attackers can rent servers in different countries, use tools developed by other people, and deliberately include misleading information.}
Hudson Rock described the evidence as supporting a relationship between the toolkit and its delivery infrastructure, not a confirmed attribution to a specific operator.

Code and Core

Why Didn’t Website Security Systems Block the Attack?

Many of the affected organizations were large companies with established security teams. So why did the malicious JavaScript still run? One major reason was that it arrived through a service the websites already trusted.

The problem with Content Security Policy

A Content Security Policy, or CSP, is a set of browser security rules.

It helps website owners control which scripts, resources, and network connections a webpage can use.

For example, a website might allow JavaScript from its own domain and a few approved third-party services. Anything outside that list can be blocked.

But in this attack, the malicious code was delivered through Optimizely’s legitimate domain. And affected websites had already approved Optimizely.

So simply trusting the source domain wasn’t enough.

Report URI reviewed the CSP configurations of affected websites and reported the following findings. The findings showed that some sites had no CSP at all. Others had policies that still allowed the malicious script to execute and transmit data. Seventeen sites reportedly allowed execution but blocked outgoing data theft. That still left room for actions such as cryptocurrency address replacement. Only 24 were reported as completely blocking the attack.

Why was unsafe-eval another concern?

Some Optimizely functionality may require a CSP setting called unsafe-eval. This setting allows JavaScript to execute code generated from strings. It can be useful for certain dynamic features, but it also weakens an important browser protection. Report URI said this was one factor that helped the attacker’s embedded backup code run. At HSBC, for example, the security policy reportedly blocked the main external payload and data exfiltration, but the embedded malicious code could still execute.

Why Couldn’t Subresource Integrity Stop the Attack?

There is another browser security mechanism called Subresource Integrity, or SRI. It works by checking a downloaded file against a known cryptographic fingerprint. If the file changes, the fingerprint no longer matches and the browser can refuse to execute it. That sounds like an ideal solution. But there is a problem with dynamic experimentation scripts.

An Optimizely snippet is designed to change whenever marketers update experiments. If a website pinned the entire file to one fixed fingerprint, legitimate experiment changes would break that check. Optimizely’s own integrity guidance therefore focuses on the stable portions of its code rather than treating the complete changing snippet as a fixed file. Payment card industry guidance also recognizes the practical limitations of applying traditional SRI checks to dynamic third-party scripts.

This creates a difficult security challenge.

The same flexibility that lets marketers change a webpage without a developer can make the underlying script harder to verify using a fixed fingerprint.

This Wasn’t the First Major Third-Party JavaScript Attack

The Optimizely incident was unusual in scale, but the underlying problem wasn’t new. Other major attacks have shown how dangerous compromised third-party JavaScript can be.

Ticketmaster UK attack in 2018

Ticketmaster UK used a third-party chatbot from a company called Inventa. Attackers compromised the chatbot’s code, which was loaded on Ticketmaster’s payment pages. The malicious code was used to steal customer payment information. The UK’s Information Commissioner’s Office later fined Ticketmaster £1.25 million. The regulator said approximately 9.4 million customers across Europe were potentially affected. It also reported that Barclays had identified 60,000 payment cards affected by known fraud.

The important point is that the attackers didn’t need to directly compromise Ticketmaster’s main payment application to reach customer information.

They exploited code supplied by a trusted third party.

Polyfill.io incident in 2024

Another example involved Polyfill.io. It was a widely used service that helped websites support older browser features. After the service changed ownership, security researchers discovered malicious code being delivered through its scripts. Security company Sansec estimated that more than 100,000 websites were affected. The US Treasury later sanctioned the company that acquired the service, Funnel. Both incidents demonstrated the same broader problem seen in the Optimizely campaign.

A website may be secure internally, but visitors can still be exposed if the external scripts it relies on become malicious

What Do Payment Security Rules Say About Third-Party Scripts?

Payment security standards have become stricter because of attacks like these.

Since March 2025, PCI DSS requirements have included stronger controls for scripts running on payment pages.

Organizations must maintain an inventory of payment-page scripts, justify their use, authorize them, and implement mechanisms to detect unauthorized changes. These controls are intended to reduce the risk of payment card skimming. But there is a limitation. The specific PCI DSS payment-page script requirements do not automatically cover every script running on a homepage, news article, or other non-payment page. That matters because a compromised experimentation tool can load on many pages outside checkout. The Optimizely incident showed that harmful JavaScript does not need to appear on a payment page to create risks for website visitors.

What Should Website Visitors Do After This Attack?

The steps depend on what someone did while visiting an affected website.

If you made a cryptocurrency transaction

Review your recent transfers, especially if you copied a wallet address while an affected webpage was open. Before sending cryptocurrency, compare the full destination address. Checking only the first and last few characters is not always enough. If you use a hardware wallet, verify the destination address on the device’s own trusted screen before confirming the transaction.

If you entered payment card details

Review your recent transactions and watch for unexpected charges. If you notice suspicious activity, contact your bank or card issuer. A payment entered through a properly isolated third-party payment field may not have been accessible to this script, but the exact risk depends on how the website implemented its checkout.

If you use an iPhone or iPad

Install the latest security updates available for your device. The exploit kits described in this incident targeted older iOS versions. Keeping the operating system updated is one of the most important ways to reduce exposure to known vulnerabilities.

What Should Website Owners and Developers Learn From This Attack?

The biggest lesson is that installing a trusted third-party tool doesn’t remove the need to monitor what that tool does. Website owners should know which external scripts are running, what permissions they have, and how those scripts change over time.

Here are the most important actions.

  1. Audit third-party JavaScript. Keep an inventory of external scripts used across the website, including analytics, chatbots, A/B testing, advertising, and marketing tools.
  2. Monitor script changes. Use tools that can detect unexpected changes in third-party JavaScript rather than checking only whether a script comes from an approved domain.
  3. Review account access. Remove unused accounts, restrict administrative permissions, and rotate credentials or tokens when compromise is suspected.
  4. Strengthen Content Security Policy. Limit where scripts can load from and where data can be sent. Avoid overly broad permissions.
  5. Avoid unsafe-eval where possible. Review whether third-party features genuinely require it and whether safer configurations are available.
  6. Enable security reporting. Use CSP reporting and other monitoring mechanisms to identify suspicious browser behavior.
  7. Review payment forms. Use properly isolated payment fields and follow applicable PCI DSS script security requirements.
  8. Prepare a third-party incident response plan. Know how to disable or replace a compromised external service quickly.

Why This Attack Matters for the Future of Website Security

Most people think of a website as one product built and controlled by one company. But modern websites work differently. A single webpage may load code from analytics providers, advertising platforms, customer support tools, personalization services, payment providers, and many other external systems.

According to the HTTP Archive figure cited in Report URI, major websites load a median of 66 external services.

Each integration adds functionality. But each also creates another dependency. When one of those services is compromised, the consequences can extend far beyond the service provider itself. That is exactly what happened with Optimizely. A malicious change inside a trusted experimentation platform reached websites associated with 178 organizations. The attacker didn’t need to replace the BBC’s homepage or break into Zoom’s internal network. The browser simply loaded a script from a trusted provider, as it was already configured to do.m And that trusted script carried malicious instructions.

Code and Core

Final Thoughts

The Optimizely cyberattack shows how a small change in a third-party tool can create a major security problem. The attackers exploited a vulnerability in a platform that many businesses relied on for website experimentation. From there, they distributed JavaScript designed to redirect cryptocurrency payments, collect card details, steal browser information, and attempt to exploit older Apple devices. Researchers detected the activity, and Optimizely said it fixed the vulnerability and secured affected accounts. But the incident exposed a wider issue that website owners cannot ignore.

Website security isn’t only about protecting your own code. It’s also about knowing what the code you trust is doing inside your visitors’ browsers.

Sources and Further Reading
This article is based on the following reports and primary sources.

World’s Largest Brands Infected with Malware via Optimizely, Report URI
Attackers Used Optimizely to A/B Test Malware on Top Websites, Krebs on Security}
Optimizely Supply Chain Attack Infects Major Brands’ Sites with Malware, BleepingComputer
Website Management Platform Optimizely Confirms Hackers Put Malicious Code into Customer Sites, The Record
Optimizely Flaw Exploited to Serve Malicious Code, Infosecurity Magazine
Optimizely Says Attacker Used a Signup Flaw to Inject Malware, Search Engine Journal
CNN, BBC, Zoom, and Other Sites Served Malicious Code in Supply-Chain Attack, Malwarebytes
OptimJack: When Trusted JavaScript Becomes a Supply Chain Attack, ReversingLabs
Optimizely Status Page

Stay Updated (Blogs)

Catch wind of the latest technologies, strategies, and information that are set to boost your business operation. We update frequently!

Looking for reliable white label services?

At Code and Core, your data is safe with top-tier encryption. For extra peace of mind, we're happy to sign an NDA to ensure full confidentiality

Hire Us
Let's Talk
  • Pay roll Basis
  • Hire Tech Pool
  • Maintenance of Existing Project
  • Fixed Price Project
  • Hourly Based
  • Something Else