0x5869697a5869…5869697d
Website Scripts Target Crypto Wallets and iOS Exploit Chains
Website-delivered scripts targeted crypto transactions and tried to load the Coruna and Darksword iOS exploit chains, selecting the payload by the iOS version detected.

Outputs
Two named iOS exploit chains referenced: Coruna and Darksword
Payload selected according to the iOS version detected during each website visit
One script targeted both cryptocurrency transactions and iPhone/iPad devices
No attacker attribution, infection count or dollar figure disclosed
No first-seen date published, so no retro-hunting window exists
Malicious code carried by ordinary websites both targeted cryptocurrency transactions and attempted to load two named iOS exploit chains, Coruna and Darksword, on the iPhones and iPads of site visitors, according to a threat report that documents the mechanics and names no operator.
The crypto targeting ran inside the browser session, not through a downloaded application or a seeded app-store listing. The report never identifies the delivery vector, and a compromised legitimate site, a malvertising buy, a typosquat and a rogue referral trail all produce the same visible symptom.
What it does establish is narrow and still consequential: the code travelled by website, and it reached both wallet-facing activity and the device in one visit. In plain terms, the exposure window is the page load rather than the app install.
How did the script choose its payload?
On iPhones and iPads the same code attempted a second-stage load of an exploit chain, and two payloads are named: Coruna and Darksword.
Rather than firing one payload at every visitor, the script selected a chain based on the iOS version detected during the visit. That points to an operator maintaining version-specific tooling and expecting to match it to the handset in front of it.
The report lists no iOS version it fingerprinted, and it does not say how reliable that detection proved. The efficiency of the selection logic therefore remains unverified, as does whether any exploit attempt succeeded on a device.
Two named chains and a version-detection step are the entirety of the technical detail on offer, and a reader cannot derive from it which devices were exposed.
What does the report not disclose?
The brief names no threat actor, group or country of origin, and it offers no count of compromised domains, infected devices or intercepted transfers, and no dollar value for the crypto activity it targeted.
Also absent:
- Threat-actor attribution of any kind
- The specific iOS versions or device models the fingerprints matched
- Numbers of compromised sites, infected devices or diverted transfers
- Indicators of compromise: domains, script URLs, wallet addresses or file hashes
- The dollar value of the transactions the scripts sought to redirect
- A first-seen date, a disclosure timeline or a patch reference
- An author, research team or vendor byline
Every one of those omissions limits what the item can support.
Why does pairing asset risk and device risk matter?
The report places two distinct monetizations in the same exposure window. A visitor who opened a flagged page on an iPhone faced both a browser-level attempt on cryptocurrency transactions and an attempt to compromise the device itself.
That combination raises the cost of a careless click beyond a phishing page, because the fallback requires no second interaction, no second download and no second visit. A single page load can carry both an asset play and a device play.
It also explains why the report treats the campaign as one incident rather than two. The crypto targeting and the exploit-chain delivery shared the same script, the same delivery surface and the same fingerprinting logic.
Skeptical readers should hold the severity line here. Reports of this class often arrive with an infection rate attached; this one arrives without one. Without a sample, a hash or a decryption key, the report's own wording that visitors' "crypto activity and devices" were at risk describes intent rather than a measured outcome, and the distance between an attempt and a completed compromise is precisely where the evidence is missing.
What should operators watch for?
With no first-seen date published, there is no stated window in which site operators can retro-hunt the injected scripts, so the practical deadline sits with site operators and mobile-security vendors to publish indicators and patched builds.
Until a sample, hash or victim count appears, the campaign is best treated as active but unmeasured. Absence of numbers is not evidence of containment.
via tokenpost.com (Source)
More from Elena Vasquez
Show full bio
Staff writer covering regulation, policy and their impact on crypto and web3 industry at Mempool Brief.
1,525 articles