Quick Take
- Researchers uncovered 188 infected website scripts linked to 178 organizations, including Ledger, in a widespread malware campaign.
- The attack exploited trusted website software to hijack crypto payments, steal card information and potentially target vulnerable iPhones.
- Despite a major cleanup, embedded malware and unanswered questions about the initial breach leave the campaign's full impact uncertain.
Ledger's website was among 178 organizations exposed to malicious scripts that could redirect cryptocurrency payments to hackers, researchers reported Monday.
Cybersecurity firm Report URI discovered the malware in 188 website scripts associated with major brands, including Ledger, BBC, CNN, Zoom, Salesforce and HSBC.
The firm examined 1,611 scripts delivered through Optimizely, a platform businesses use to test and personalize their websites. Attackers had inserted malicious instructions into customer-specific JavaScript, allowing infected code to reach visitors through a service those websites already trusted.
The attack exposed a vulnerability for cryptocurrency users: hackers could redirect funds without stealing private keys or compromising wallets.
Ledger's inclusion in the affected websites does not establish that its hardware wallets, firmware, or signing infrastructure were breached.
How hackers hijacked crypto payments without stealing private keys
The malicious code was designed to identify crypto addresses displayed on websites and replace them with destinations controlled by attackers.
Researchers at Hudson Rock found that the malware could alter wallet addresses in visible text, input fields, links, and clipboard operations. It supported multiple blockchain networks, including Bitcoin, Ethereum, Solana, Tron, and Litecoin.
For example, a user copying a legitimate Bitcoin payment address from an infected page could unknowingly paste an attacker-controlled address into their wallet application.
If the user then approved the transaction without verifying the destination, the funds would go to the attacker.
The technique circumvents an important layer of cryptocurrency security because the wallet can function exactly as intended while authorizing a payment to the wrong recipient.
Hudson Rock found no evidence that the recovered malware directly extracted wallet seed phrases or initiated unauthorized blockchain transactions through wallet-signing functions.
A trusted script can change the payment destination
Researchers described code that could alter a displayed or copied crypto address before a transfer, without needing the private key.
Why a host block may leave address swapping available
Illustrative flow: the remote payload reaches the page. The recovered snippet also contains a fallback.
Illustrative flow for embedded-fallback variants: a host block may stop the fetch while local code remains available, subject to browser policy.
Destination substitution can divert a user-authorized payment without the malware signing it.
Presence of poisoned code does not prove execution or a completed transfer. The fallback concerns two recovered variants and remains subject to browser policy. No seed extraction or wallet-signing calls were identified in the recovered samples; this is not a blanket security assurance.
However, the attack appears to have generated financial proceeds.
Report URI identified approximately $7,040 received by attacker-linked crypto addresses as of 19:25 UTC on Nov. 16. Bitcoin accounted for roughly $5,380, while additional transactions involved Litecoin, Tron, Ethereum, and BNB Smart Chain.
Beyond cryptocurrency theft, the malware included additional capabilities to intercept payment card information, collect browser-accessible cookies, and capture screenshots.
Researchers also identified code associated with the Coruna and DarkSword iPhone exploit kits, potentially exposing visitors using vulnerable Apple devices to additional attacks.
These capabilities expanded the threat beyond crypto holders to customers interacting with infected websites.
Report URI said a large-scale cleanup reduced the number of infected scripts in its sample from 169 at 20:56 UTC to just one at 21:52 UTC on Nov. 16. Ledger's affected script was reportedly cleaned at approximately 20:41 UTC.
However, the researchers also discovered embedded copies of the malicious code, meaning blocking the attackers' remote servers alone might not eliminate every infection.
The central mystery is how the attackers got permission to modify scripts for so many unrelated organizations.
The researchers have not established whether the intrusion originated from compromised credentials, an Optimizely system, or another access route.
At publication, the investigation had not established the number of affected visitors or the full extent of financial losses, leaving the attackers' initial access method and the campaign's ultimate impact unresolved.


