Decision RecordActivePublished without chair review
CRT-2026-042525 Oct 2026AFTERNOON EDITIONDaily Roundtable
Enterprise response to plausible DarkSword exposure
Update supported devices, restrict or isolate devices that cannot receive current security fixes, reconstruct browser and advertising-delivery timelines, preserve suspect devices and network evidence, and rotate sessions or credentials from a clean device when compromise is credible. Do not center the response on mass application removal.
Current public guidance · the full record
What to do now
At a glanceThe edition's authoritative action board carries no action for this record's subjects — no What to do now guidance.
Why now
Partially supportedThe material available on October 25, 2026 describes observed DarkSword and Coruna delivery through compromised advertising or analytics infrastructure and a malicious Instagram ad, while the reviewed DarkSword reporting describes a no-click web exploit chain capable of kernel-level access and post-exploitation access to selected Keychain, application-container, and wallet data.
Those consequences justify immediate updating, isolation, exposure reconstruction, and evidence preservation when exposure is plausible. The same evidence does not justify delaying action for a complete version matrix or diverting the response into mass application removal.
Because this is a historical replay bounded to October 25, 2026, later fixes or findings are not verified here.
Who is affected
Partially supportedOperators of supported iOS devices with plausible exposure through compromised websites, advertising or analytics infrastructure, redirects, a malicious Instagram ad, or fake Apple ID pages need to update those devices and reconstruct exposure.
Operators of iOS devices that cannot receive current security fixes need to restrict or isolate them, although the packet does not identify a complete version or iPhone-model list.
Users of plausibly compromised iPhones face possible exposure of selected Keychain data, application-container data, wallet-related data, and authenticated sessions.
Security teams managing high-consequence iPhones need to preserve device and network evidence and escalate when browser, network, device, authentication, or wallet indicators correlate.
Owners of installed applications are not, on the reviewed evidence, a distinct infection population merely because DarkSword can enumerate applications or access their containers after exploitation.
What supports this
SupportedSupport: The mobile-security analysis describes DarkSword as a six-vulnerability, no-click web exploit chain capable of reaching kernel-level access and reports delivery through compromised advertising or analytics infrastructure and a malicious Instagram ad.
Support: The follow-up analysis identifies compromised legitimate websites, redirects, and fake Apple ID pages as supported delivery paths and reports no credible evidence of trojanized App Store applications, enterprise-signed applications, or malicious application updates carrying DarkSword.
Support: The same reporting describes commands for enumerating installed applications and extracting selected Keychain, application-container, and wallet data, which supports preserving evidence and protecting sessions and credentials after credible compromise.
Support: The evidence auditor found the packet consistently supports updating supported devices, restricting devices without current security fixes, reconstructing web and advertising exposure, preserving device and network evidence, and rotating credentials from a clean device.
Evidence gap: A second audit found no complete device-model, iOS-version, exploit-chain, and Apple-fix mapping, limiting precise population claims.
How the Roundtable reached this
Partially supportedThe mobile-security specialist separated DarkSword delivery, exploitation, and post-exploitation collection.
She found support for a no-click web exploit chain delivered through compromised websites, advertising or analytics infrastructure, redirects, a malicious Instagram ad, and fake Apple ID pages, but not for delivery through trojanized App Store applications, enterprise-signed applications, or malicious application updates.
The moderator adopted that distinction and rejected mass application removal as the primary response.
The decision scout translated the finding into operational steps: update supported iOS devices, restrict devices without current security fixes, reconstruct exposure, preserve evidence, and rotate sessions or credentials from a clean device when compromise is credible.
The evidence auditor supported those steps while identifying the missing device-model, iOS-version, exploit-chain, and Apple-fix mapping. The boundary reviewer resolved the scope concern by avoiding a narrowly defined high-risk population.
A record search found no matching prior decision, and the arbiter selected a new operational decision record.
Positions are generated by AI specialist personas and chaired by Halil Öztürkci.
Panel composition
- Scout (AI panel role)Scout identified 10 candidate signals.
- Linker (AI panel role)Linker evaluated 10 relation judgments.
- Evidence Auditor (AI panel role)Evidence Auditor recorded 23 evidence signals; 13 gaps.
- Prediction Steward (AI panel role)Prediction Steward accepted 0 predictions and rejected 2 claims.
- Boundary Reviewer (AI panel role)Boundary Reviewer recorded 17 public/private findings.
- Arbiter (AI panel role)Arbiter produced 10 decision envelopes.
Key disagreement
Scout (AI panel role)
Observed delivery does not establish widespread successful exploitation, universal module success, durable persistence, or quantified wallet loss. Application-based delivery is not proven impossible, but it is unsupported by the reviewed evidence.
Arbiter outcome
Arbiter outcome: new decision record. The operational response is well supported, no matching record was found, and unresolved device-version precision can be handled by avoiding a narrowly defined affected population.
Candidates considered
Considered 10 candidates · opened 1 · 9 not opened (9 other)
Considered, not opened
Sign in to preview Considered-Not-Opened entries (moves to Pro at launch).
Sign in to preview practitioner entries.
What is uncertain
ConflictingThe exact iOS versions and iPhone models exposed to each DarkSword exploit chain remain unresolved, as does the correspondence between those chains and Apple fixed releases.
The reviewed material does not establish widespread successful exploitation, success of every module, durable persistence, or quantified wallet loss. Application-carried delivery is not proven impossible, but the reviewed evidence does not support it.
These limits affect population sizing and incident conclusions, not the broad containment and evidence-preservation steps.
What evidence is missing
ConflictingThe available evidence lacks an authoritative mapping of DarkSword exploit chains to specific iOS versions, iPhone models, Apple fixes, and fixed releases.
It also does not quantify successful exploitation, module reliability, durable persistence, or wallet-related loss. Direct vendor advisories and the underlying public reports are not attached in the bounded packet.
Until those materials are available, do not label a precise model or version range as high risk, and do not merge older Coruna coverage into DarkSword coverage.
What would change this
Partially supportedAuthoritative evidence that identifies exact affected iOS versions, iPhone models, and Apple fixed releases would replace the broad supported-versus-no-current-fix split with a precise target population.
Signed-application, malicious-update, or filesystem evidence demonstrating application-carried DarkSword infection would change the delivery model and could make targeted application removal part of the primary response.
Evidence disproving the reported web and advertising delivery paths would also require rebuilding the exposure-reconstruction plan.
What to watch next
Partially supportedWatch for authoritative Apple or research mappings that connect DarkSword exploit chains to specific iOS versions, iPhone models, and fixed releases; use that mapping to narrow patching and isolation priorities.
Correlate browser and advertising history with device, network, authentication, Keychain, application-container, and wallet indicators; escalate to specialist mobile forensics when multiple evidence sources align on a high-consequence device.
Watch for signed-application, application-update, or filesystem evidence showing that an application carried DarkSword; if found, expand the response beyond web-delivery reconstruction and reassess application removal.
Evidence basis
DarkSword should be treated as a web-delivered exploit chain, not as evidence that compromised applications were already installed on victims’ iPhones. The reporting Nadia reviewed supports delivery through compromised legitimate websites, …
Summary: The room put the PHP flaws first: AWS MadPot observed probing of CVE-2026-109618 and CVE-2026-109619 before patches were available, while Datadog reportedly reproduced exploitation before application code executes. Executed LiteLLM…
Public value history
- 25 Oct 2026Initial public guidanceCurrent guidance
Created the first public value version for this Decision Record.
Source RoundtableAfternoon roundtableConvened 25 Oct 2026Methodology
How the panel reaches a Public Decision Record.