Intelligence Brief: Optimizely Script Exploitation Injects Card Skimmer and iOS Exploits Targeting CNN, BBC

Sovereign Geopolitical Intelligence &
Situational Awareness Terminal
[SYSTEM STATUS: OPERATIONAL]
[INGESTION RATE: — briefs/day]
[THREAT LEVEL: ELEVATED]

◈ Source Credibility Index

Multi-source assessment (4 sources)(ibtimes.co.uk)3/5 — Generally ReliableNATO C/3 — Fairly Reliable / Possibly True

1. BLUF (Bottom Line Up Front)

An unidentified attacker used a flaw in Optimizely Web Experimentation to write malicious JavaScript into customer projects on 16 November 2026. The code ran on sites of at least 178 organizations, including CNN, the BBC, Zoom, Salesforce, Equifax and IKEA. It swapped cryptocurrency wallet addresses, skimmed card details and tried to load iOS exploit kits. The best-supported explanation is a single, mainly financially motivated actor abusing a platform-side access weakness. Report URI counted roughly $6,600 to $7,040 in attacker-wallet receipts and no confirmed iPhone takeovers. Overall confidence: probably (about 66%). The root cause and the actor's full objective are not established.

2. Key Judgments — Optimizely Supply-Chain Script Injection

  1. It is highly likely that malicious code was served through Optimizely's legitimate infrastructure. Optimizely confirmed an unauthorized party exploited a vulnerability, and Report URI, Cynet and Hudson Rock (via Reuters) each describe the code reaching visitors. The Hacker News reports that snippets still came from cdn.optimizely.com under the vendor's certificate, so a script allowlist would not have blocked them.
  2. It is likely the attacker was financially motivated and opportunistic. Reported functions were wallet-address replacement, a card skimmer added at 13:43 UTC, and cookie and screenshot exfiltration. Measured proceeds were small: about $7,040 (Report URI) or $6,600 (StepSecurity). StepSecurity noted that two addresses predate the injections and that the chain does not show which transfers came from victims.
  3. The iOS component is a real capability but shows no demonstrated impact. Coruna (iOS 13.0–17.2.1) and DarkSword (iOS 18.4–18.7) were loaded conditionally. Cynet retrieved the second stage only through an emulated iPhone browser and called it "not a successful compromise of a physical iPhone or iPad." Absence of observed compromise is not proof of none.
  4. The 178-organization figure is a floor. Report URI checked 1,611 snippets, found code in 188 belonging to 178 organizations, and said the true number is likely higher. A site serving the code is not necessarily compromised internally.
  5. Remediation looks complete but its durability is unproven. Smartsheet's script was cleaned and then re-infected with a hidden version, which indicates the attacker kept access at that point. Whether access has been fully removed is not reported.

3. Analysis of Competing Hypotheses (ACH)

Hypothesis Supporting Evidence Contradicting Evidence Evidence Gaps Probability
H-A: A single, mainly financially motivated actor exploited a platform-side flaw in Optimizely to mass-write Project JavaScript. The iOS kits were an opportunistic add-on. Optimizely confirmed exploitation of a vulnerability. The payload targets wallets and cards, with measured receipts near $7,000. First injection came at 04:21 UTC in the BBC snippet, mass cleanup at about 20:40 UTC, and a skimmer was added mid-campaign at 13:43 UTC. The same loader family went through three delivery variants. The iOS chain was a substantial investment for a roughly $7,000 return. The 60-second reload retry is noisy. The re-infection of Smartsheet suggests persistent access, which a one-off flaw fits less well. Root-cause detail from Optimizely. Attacker identity. Whether one actor or several operated. 47%
H-B: Access came mainly through account-level abuse (the signup route, sessions, API tokens) rather than a flaw in the script-delivery code. "Vulnerability" is the vendor's umbrella term. IBTimes reports Optimizely said the flaw was in its Experimentation account signup process and that it terminated sessions and revoked access tokens. Report URI's attacker wrote to "so many unrelated customer projects." Report URI advises reviewing users, tokens and integrations. The Hacker News says Optimizely's notice does not explain how the flaw worked. The signup attribution appears in one outlet only. H-B overlaps with H-A and may be a refinement of it rather than a rival. Optimizely's full post-incident statement. Count of projects altered. Token and session logs. 26%
H-C: Compromising iOS devices was the primary objective, and crypto theft was secondary or cover. Two documented exploit kits were wired in. A separate iOS 26 chain collected device details. A Coruna iframe was pushed to every iOS 18.6-or-earlier visit between 17:13 and 17:57 UTC. The same actor's injector had run on other vendors' scripts since late last week and on Turkish sites from 17 September. No device takeover has been reported. The kits' own comments mark iOS 18.7.3+ and 26.3+ as patched. Wallet and card theft features were built into the main payload. Endpoint telemetry from real devices. Apple or Google confirmation of exploitation in the wild. Post-exploit payload analysis. 21%
H-D (Maskirovka / Strategic Deception): The apparent signal is a deliberate disinformation, fabrication, or denial-and-deception operation designed to shape perception or mask a different course of action. HSBC told Reuters that unauthorized scripts were not active on its websites, although Report URI lists it. One outlet's headline names Fox News, which the other sources do not. Optimizely's initial notice was sparse. The Hacker News notes the wallet receipts include addresses used before the injections. Report URI, Cynet, StepSecurity, Hudson Rock, Reuters and Optimizely independently describe the event. Report URI published hashes, timestamps and the full organization list. Equifax and Salesforce took protective action. Independent verification of HSBC's scripts. Archived snippet copies reviewed by CERTs. 3%

ACH Assessment: H-A is best supported because it fits Optimizely's confirmation, the payload's functions and the observed timing with the fewest extra assumptions. H-B is a plausible refinement of H-A rather than a clean rival. H-C cannot be excluded because the iOS tooling is expensive relative to the crypto proceeds. The one flagged contradiction is a naming difference between two claim summaries that attribute findings to Helme, Report URI and Hudson Rock. It does not contradict the core facts. The real disagreements are narrower: the root cause, the cleanup end time, total losses, and whether HSBC's site was affected. Together they lower confidence in details but not in the core event.

4. Key Assumption Check (KAC)

  • Critical Assumptions:
    • Report URI's crawl-based counts are broadly accurate. If it over-attributed snippets, the 178 figure falls; the sample covers only snippets it found.
    • Wallet receipts approximate victim losses. StepSecurity says the chain does not isolate victim transfers, so actual theft could be lower or higher.
    • The campaign was a single operation. If the earlier webshell[.]lol, Kiss, StorifyMe and Minute Media injections belong to a different actor, the "same actor" linkage fails.
    • Cleanup is complete. The Smartsheet re-infection and the TalkTalk remnant at 21:52 UTC show the attacker's access was not removed in one step, and reinfection would extend the exposure window.
  • Information Gaps: Optimizely has not disclosed how the flaw worked, how many projects were changed, when the fix went in, or whether the attacker retains access. Closing this requires a vendor post-incident report. The number of cards or devices actually harmed is unknown; issuer fraud data and iOS telemetry would address it. No attribution is reported. The cleanup end time differs: The Hacker News says only TalkTalk remained at 21:52 UTC, and IBTimes says all 188 were clean by 11:11pm UTC. Clarification would come from Report URI's timestamped snapshots.
  • Bias & Deception Risks: Three of the four outlets draw on Report URI and Reuters, so apparent agreement partly reflects a shared origin and is not full independence. The consumer-advice pieces use broad language ("hacked websites") that overstates compromise. One loosely edited outlet lists Fox News as affected, which no other source supports. Companies have reputational reason to minimize: HSBC's denial and Equifax's "no evidence of compromise" are limited statements and not proof the code never ran. Report URI is a vendor with a commercial interest in script-monitoring visibility. No clear adversary deception indicators appear.

5. Implications and Strategic Risks — Optimizely and Downstream Customers

Because one account-write weakness reached many unrelated customers at once, this is a shared-dependency failure. The attacker had already used webshell[.]lol through other vendors' scripts, so further third-party script vendors may carry similar exposure. If persistence is confirmed, a second wave could carry a different payload. Reuters has not independently determined the extent of any successful attacks.

Cyber / Information Space — Third-Party JavaScript Supply Chain

Report URI says Content Security Policy allowlisting cdn.optimizely.com would not have helped, because the code came from that host. Strict-dynamic and unsafe-eval let the payload run on some sites. The skimmer cannot read cross-origin payment iframes such as Stripe or Adyen hosted fields, so sites using them were less exposed to card theft. Expect scrutiny of write-access controls on experimentation, tag-management and analytics platforms.

Economic / Social — Affected Brands and Consumers

Salesforce disabled the scripts, Equifax stopped using Optimizely, and IKEA is assessing effects on customers. Card issuers may see delayed fraud from cards entered on 16 November. Report URI says it could not see how many cards were exposed. Some customers may switch vendors, and breach-notification questions arise where card data reached the attacker's server.

Security / Counter-Terrorism — iOS Exploit Kit Proliferation

The delivery of Coruna and DarkSword through a commercial web platform shows these kits being used outside targeted operations. Devices on iOS 18.7.3+ or 26.3+ are marked patched by the kits' own comments. Older-version devices that visited affected sites on 16 November are the unresolved exposure.

6. Recommendations and Outlook

  • Immediate Actions (0–30 days): Track Optimizely's incident page for a root-cause statement, scope numbers and any statement on attacker access. Watch Report URI's published list for additions beyond 178 and for reinfection of cleaned snippets. Monitor attacker-wallet inflows, the 1688360[.]cc, webshell[.]lol and .icu domains, and card-fraud reporting tied to 16 November. Check organizational Optimizely projects for 1688360, atob( or unfamiliar Project JavaScript, as Report URI advises. Review which accounts and tokens can write to those projects.
  • Medium-Term Posture (1–12 months): Assess whether script-integrity controls, strict script-src and connect-src policies, and write-access audits are in place for all third-party script providers. Watch for regulator or card-scheme inquiries and for litigation naming Optimizely or affected brands. Track whether the same actor reappears through other script vendors, and whether Apple or Google publish further exploit-in-the-wild confirmation.
  • Scenario Outlook:
    • Most likely: Optimizely publishes a limited post-incident report, losses stay in the low five figures, no confirmed iPhone compromises appear, and the actor shifts to other script vendors. Trigger: continued activity on the other domains.
    • Best: Access is fully removed with no reinfection, and the vendor shares logs with CERTs. Trigger: a clean follow-up crawl and a root-cause disclosure.
    • Worst: The attacker retains hidden access and returns with a heavier payload, or evidence emerges of real device compromise or large card exposure. Trigger: renewed injections, fraud spikes, or device-compromise reports.

7. Key Individuals and Entities

Name Role / Affiliation Relevance to Assessment
Unidentified attacker Operator of the injection campaign (called OPTIMJACK by some researchers) No attribution reported. Also tied by Report URI to earlier webshell[.]lol injections.
Optimizely Web Experimentation vendor Confirmed the exploit, removed the code and revoked tokens. Root cause and scope are not disclosed in full.
Scott Helme / Report URI Script-monitoring firm and its founder Primary source for the 178-organization count, timeline and payload details.
Cynet Endpoint security vendor Traced the loader to cdn.optimizely.com and captured the code on CNN and Zoom. Its emulated-browser test did not show a real iPhone compromise.
StepSecurity Security analysis firm Payload analysis and the $6,600 wallet figure.
Hudson Rock Cybersecurity firm Told Reuters it found affected sites served malicious code.
Reuters News agency Source for company statements. Reported it could not independently determine the extent of successful attacks.
CNN, BBC, Zoom Affected sites Named across sources. The BBC snippet carried the first injection.
Salesforce, Equifax, IKEA, HSBC Affected organizations Salesforce disabled the scripts, Equifax stopped using Optimizely, IKEA is investigating, and HSBC said unauthorized scripts were not active on its sites.
Smartsheet, TalkTalk Affected customers Smartsheet's snippet was re-infected after cleanup. TalkTalk was the last snippet still carrying code at 21:52 UTC.



Explore more: Cybersecurity Briefs · Daily Summary · Support us

WorldWideWatchers · Intelligence Assessment
Source Verification & Governance Report

2026-11-18 11:17:03 UTC
595be098

Source Reliability
3
Generally Reliable
Source Credibility Index

NATO C · Fairly Reliable
4 source(s) · 4 domain(s)

Information Credibility
PASS
58% faithful
AI faithfulness check

NATO 3 · Possibly True
Corroboration: 82% (STRONG) · Conflicts: 1 · MEDIUM

Governance Decision
Cleared
✓ YES Publication
✓ YES Dissemination
✓ Cleared Analyst review

Corroborating Sources
Source SCI Role
economictimes_indiatimes 3 SOURCE_DOCUMENT
The Hacker News 4 SOURCE_DOCUMENT
headtopics 3 SOURCE_DOCUMENT
ibtimes 2 SOURCE_DOCUMENT
⚠ Detected Conflicts (1)
  • NLI CONTRADICTION (74%): NLI contradiction=0.740 ≥ threshold=0.65. Claim A: "Hackers, Optimizely, cybersecurity researcher Scott Helme, Hudson Rock, CNN, Zoom, IKEA, HSBC, Equ
Generated by WorldWideWatchers Intelligence Pipeline · 2026-11-18 11:17:03 UTC · Machine-generated assessment — subject to analyst review before operational use.