Skip to navigation

Threat reportThreat ActorTL-2026-1417

Iran's AI-Enhanced Asymmetric Playbook: State Actors Integrate AI Across Cyber, Influence, and Military Operations (2026 Conflict)

highACTIVE

Recorded Future documents Iranian state-linked actors — IRGC-affiliated groups, MOIS, APT42/GreenBravo, MuddyWater/GreenGolf, APT34/OilRig, RedKitten, Dust Specter, Nimbus Manticore, plus hacktivist personas CyberAv3ngers, Cyber Isnaad Front, and Ababil of Minab — integrating AI (Gemini, ChatGPT) into cyber operations, influence campaigns, military systems, and domestic surveillance during the January-June 2026 conflict with the US and Israel.

Key facts

IdentifiersTL-2026-1417
Also known asOperation Olalampo, Operation Epic Fury, Operation Roaring Lion
SeverityHIGH
Affected productsVarious Microsoft Office (macro-enabled documents); Various Industrial Control Systems / PLCs (industrial refrigeration, energy, transportation, maritime); Various Aviation, defense, aerospace, and telecom sector enterprise networks (US, Europe, Middle East)
AttributionAPT42, GreenBravo (medium confidence)
Nation-state nexusIran
MalwareCHAR, GHOSTFORM, GhostBackDoor, GhostFetch, HTTP_VIP, MiniFast, MiniJunk, SPLITDROP, SloppyMIO, TWINTALK, TWINTASK
ToolsAnyDesk
MITRE ATT&CK techniquesT1005 Data from Local System, T1027 Obfuscated Files or Information, T1036 Masquerading, T1041 Exfiltration Over C2 Channel, T1053 Scheduled Task/Job, T1057 Process Discovery, T1059 Command and Scripting Interpreter, T1071 Application Layer Protocol, T1082 System Information Discovery, T1102 Web Service, T1105 Ingress Tool Transfer, T1119 Automated Collection and 18 more in the attack flow
MotivationESPIONAGE
Target sectorsgovernment administration, defense, aerospace, aviation, telecoms, energy, transport, maritime, criticalinfrastructure, humanrightsngo, civil society, technology
Target regionsNorth America, Middle East, Europe, Gulf States, iraq, israel, iran
Detection rules9 (1 shown free)
Sources18 cited sources

Detection coverage

As of 2026-07-16, Threadlinqs Intelligence publishes 9 detection rule(s) for TL-2026-1417 across Splunk SPL, Microsoft KQL and Sigma, covering 30 indicator(s) of compromise. The whole corpus is readable without an account. Every threat comes with one complete detection in Splunk SPL, Microsoft KQL and Sigma, free for everyone; a free account adds daily TL Creds to unlock the rest. Paid plans add a larger daily allowance, correlation and the MCP server. Threadlinqs MCP server · View plans.

Free detection: Office Macro Spawning Encoded PowerShell/CMD Child (Operation Olalampo Initial Access)

Detects WINWORD.EXE/EXCEL.EXE spawning powershell.exe or cmd.exe with encoded/hidden-window arguments, matching MuddyWater/GreenGolf's Operation Olalampo phishing-document macro chain that drops the GhostFetch downloader.

  • Languages: Splunk SPL, Microsoft KQL, Sigma
  • ATT&CK: T1566.001, T1059.001, T1059.005
  • Data sources: sysmon, windows_security, microsoft_defender_for_endpoint
  • Severity: high
  • Confidence: high

Splunk SPL

index=sysmon sourcetype="XmlWinEventLog:Microsoft-Windows-Sysmon/Operational" EventCode=1
| eval ParentImage=lower(ParentImage), Image=lower(Image)
| where (like(ParentImage, "%\\winword.exe") OR like(ParentImage, "%\\excel.exe"))
  AND (like(Image, "%\\powershell.exe") OR like(Image, "%\\cmd.exe"))
  AND (like(CommandLine, "%-enc%") OR like(CommandLine, "%-nop%") OR like(CommandLine, "%-w hidden%") OR like(CommandLine, "%-windowstyle hidden%"))
| stats count min(_time) as first_seen max(_time) as last_seen by ComputerName, User, ParentImage, Image, CommandLine, ParentProcessId, ProcessId
| table first_seen last_seen ComputerName User ParentImage Image CommandLine
| sort -last_seen

Microsoft KQL

DeviceProcessEvents
| where Timestamp > ago(24h)
| where InitiatingProcessFileName in~ ("WINWORD.EXE", "EXCEL.EXE")
| where FileName in~ ("powershell.exe", "cmd.exe")
| where ProcessCommandLine has_any ("-enc", "-nop", "-w hidden", "-windowstyle hidden", "EncodedCommand")
| project Timestamp, DeviceName, AccountName, InitiatingProcessFileName, FileName, ProcessCommandLine, InitiatingProcessId, ProcessId
| order by Timestamp desc

Sigma

title: Office Document Spawning Encoded PowerShell/CMD Child (Operation Olalampo)
id: 3f1a9c20-1417-4a01-9e01-olalampospl01
status: experimental
description: Detects Office applications (WINWORD.EXE/EXCEL.EXE) spawning powershell.exe or cmd.exe with encoded or hidden-window arguments, consistent with the MuddyWater/GreenGolf Operation Olalampo macro-to-GhostFetch chain.
references:
    - https://hivepro.com/threat-advisory/operation-olalampo-muddywater-expanding-campaign-across-mena/
    - https://thehackernews.com/2026/02/muddywater-targets-mena-organizations.html
author: Threadlinqs Team
date: 2026-07-16
tags:
    - attack.initial-access
    - attack.t1566.001
    - attack.execution
    - attack.t1059.001
    - attack.t1059.005
    - attack.group.muddywater
logsource:
    product: windows
    category: process_creation
detection:
    selection_parent:
        ParentImage|endswith:
            - '\WINWORD.EXE'
            - '\EXCEL.EXE'
    selection_child:
        Image|endswith:
            - '\powershell.exe'
            - '\cmd.exe'
    selection_args:
        CommandLine|contains:
            - '-enc'
            - '-nop'
            - '-w hidden'
            - '-windowstyle hidden'
    condition: selection_parent and selection_child and selection_args
falsepositives:
    - Legitimate mail-merge or macro-driven automation scripts using PowerShell (rare, should be allowlisted by hash/path)
level: high

Known false positives:

  • Legitimate enterprise macro automation that shells out to PowerShell for approved mail-merge or document-generation workflows

Tuning notes: Baseline known internal macro-automation tools by hash/publisher and exclude them; tighten by requiring both -enc AND -w hidden together to reduce noise from legitimate scripted deployments.

8 more detection rules (Splunk SPL, Microsoft KQL, Sigma) · unlock with TL Creds. Sign up free to unlock the detection rules or query them with the Threadlinqs MCP server (Purple)
  • GhostFetch-style Host Profiling Recon Following Office Macro Execution · unlock with TL Creds
  • AnyDesk Deployment Shortly After Native Downloader Execution (HTTP_VIP Hands-on-Keyboard Staging) · unlock with TL Creds
  • Non-Browser Process Repeatedly Fetching GitHub Raw Content or Google Drive Direct-Download Links (SloppyMIO Staging) · unlock with TL Creds
  • Sigma: Office Macro to Encoded PowerShell/CMD Child Process (Operation Olalampo) · unlock with TL Creds
  • Sigma: Unsigned 64-bit PE DLL Masquerading as Chrome Component (MiniFast) · unlock with TL Creds
  • Sigma: Password-Protected Forensic-Themed Archive Delivered to NGO/Activist Mailbox (RedKitten Shock-Lure) · unlock with TL Creds
  • GHOSTFORM Fileless PowerShell Execution From Remote-Fetched Script (No Disk Write) · unlock with TL Creds
  • Outbound Telegram Bot API C2 Traffic Outside Legitimate Telegram Client (CHAR/SloppyMIO) · unlock with TL Creds

Indicators of compromise

As of 2026-07-16, Threadlinqs Intelligence tracks 30 indicators of compromise for TL-2026-1417.

30 indicators of compromise · unlock with TL Creds. Compare plans

Technical analysis

AI functioned as a force multiplier accelerating existing tradecraft (spearphishing, malware development, ICS reconnaissance, propaganda) rather than creating breakthrough capabilities.

During the January-June 2026 US/Israel-Iran conflict (Operation Epic Fury/Roaring Lion, February 28 airstrikes, through the June 17 collapse of a US-Iran memorandum of understanding), Recorded Future assessed that Iranian state and IRGC-affiliated actors integrated generative AI tools across cyber operations, information warfare, military systems, and domestic surveillance. The clearest AI impact was in information warfare, where large language models and image generators (including Freepik.com) enabled rapid, culturally-nuanced propaganda production distributed across Instagram, Bluesky, TikTok, and X/Twitter, including a documented case of AI-generated Lego-themed propaganda video content.

On the cyber front, multiple Iranian APT clusters independently adopted AI-assisted malware development, evidenced by AI-generated code artifacts (excessive defensive/error-handling logic, embedded emoji/unicode patterns, and placeholder hardcoded seed values such as 0xABCDEF) discovered in malware from Dust Specter (TWINTALK/GHOSTFORM), Nimbus Manticore (MiniFast), and RedKitten-linked SloppyMIO. MuddyWater (GreenGolf) launched Operation Olalampo on January 26, 2026, targeting MENA organizations via phishing documents with malicious macros dropping the GhostFetch downloader, GhostBackDoor implant, HTTP_VIP native downloader, and the Rust-based Telegram-controlled CHAR backdoor. RedKitten (newly identified January 2026, overlapping Yellow Liderc/Imperial Kitten TTPs) targeted Iranian human-rights NGOs and protest documentarians with 'Tehran Forensic Medical Files' shock-lure archives delivering the SloppyMIO implant, which used GitHub and Google Drive for staging and Telegram for C2. Nimbus Manticore resurfaced during Operation Epic Fury (from February 28, 2026) targeting US, European, and Middle Eastern aviation and software/defense/aerospace/telecom organizations with the AI-assisted MiniFast backdoor (64-bit Windows PE DLL impersonating Chrome), delivered via phishing and, from April 2026, SEO-poisoned fake SQL Developer installer pages. Dust Specter (APT34/OilRig-linked, medium-to-high confidence Iran-nexus) targeted Iraqi government officials via compromised Iraqi government infrastructure using two attack chains: SPLITDROP dropper with TWINTASK/TWINTALK backdoors, and the consolidated in-memory GHOSTFORM RAT executing PowerShell retrieved from C2 without writing to disk.

Separately, Recorded Future documented an ICS/OT reconnaissance capability enabling analysts to move from expressed intent to a list of accessible US industrial control system devices within five minutes using AI-assisted search/enumeration, targeting energy, transportation, aviation, and maritime/shipping logistics sectors, and industrial refrigeration/PLC systems. Hacktivist proxy personas amplified the campaign: CyberAv3ngers (historically linked to IRGC Cyber Electronic Command and prior Unitronics PLC defacements) continued ICS-themed messaging; Cyber Isnaad Front (FDD-assessed likely Iranian proxy) claimed breaches of over 160 Israeli telecom data centers, exfiltration of 5TB from a national fuel logistics provider, and compromise of Israeli logistics/tracking and government/military communications systems, though independent verification found many claims overstated; Ababil of Minab (MOIS-linked) conducted parallel hacktivist activity. Iran also received external support: Russia supplied Shahed-236/MS001 drone hardware with Nvidia Jetson Orin AI modules, tactical/flight-planning tradecraft, satellite imagery, and FindFace facial-recognition technology; China supplied surveillance infrastructure (Huawei/ZTE deep packet inspection, Hikvision/Tiandy cameras) and geospatial-AI target-identification tooling (MizarVision), reinforcing a domestic surveillance apparatus originally built for hijab-law enforcement and expanded during an 88-day nationwide internet blackout (beginning January 8, 2026) to suppress protests.

Recorded Future's overall assessment: AI 'almost certainly enhanced' Iran's asymmetric playbook by accelerating existing tradecraft velocity and scale, but had 'not fundamentally altered' Iranian strategic doctrine; confirmed operational integration of AI into live military systems (beyond drone guidance hardware) remains unconfirmed.

MITRE ATT&CK techniques by tactic:

Collection

T1005 Data from Local System; T1119 Automated Collection

Defense Evasion

T1027 Obfuscated Files or Information; T1036 Masquerading; T1140 Deobfuscate/Decode Files or Information; T1497 Virtualization/Sandbox Evasion; T1620 Reflective Code Loading

Exfiltration

T1041 Exfiltration Over C2 Channel

Persistence

T1053 Scheduled Task/Job; T1543 Create or Modify System Process

Discovery

T1057 Process Discovery; T1082 System Information Discovery; T1518 Software Discovery

Execution

T1059 Command and Scripting Interpreter; T1204 User Execution

Command and Control

T1071 Application Layer Protocol; T1102 Web Service; T1105 Ingress Tool Transfer; T1219 Remote Access Tools

Initial Access

T1189 Drive-by Compromise; T1199 Trusted Relationship; T1566 Phishing

Impact

T1485 Data Destruction; T1491 Defacement

Resource Development

T1583 Acquire Infrastructure; T1584 Compromise Infrastructure; T1585 Establish Accounts; T1587 Develop Capabilities

Reconnaissance

T1593 Search Open Websites/Domains; T1596 Search Open Technical Databases

Remediation

Immediate actions

  • Block delivery of password-protected 7z archives from unsolicited email in high-risk NGO/activist environments
  • Disable Office macro execution by default; alert on VBA macro-spawned child processes (cmd.exe, powershell.exe, mshta.exe)
  • Block outbound Telegram Bot API traffic from endpoints where not business-justified
  • Monitor for AnyDesk deployment from unexpected parent processes following native downloader execution
  • Restrict/monitor use of GitHub raw content and Google Drive direct-download URLs as executable staging channels on managed endpoints
  • Segment ICS/OT networks from internet-reachable IT assets and restrict remote access to PLCs and industrial refrigeration controllers
  • Alert on unsigned or newly-compiled 64-bit PE DLLs masquerading as Chrome components

Workarounds

  • Enforce application allowlisting to prevent execution of downloader-staged secondary payloads in memory
  • Require out-of-band verification for unsolicited 'casualty list' or forensic-document-themed archives before opening

Longer-term hardening

  • Deploy EDR with behavioral/in-memory detection for fileless PowerShell execution retrieved from C2 (GHOSTFORM-style tradecraft)
  • Establish continuous ICS/OT asset discovery and exposure monitoring to counter AI-accelerated adversary reconnaissance
  • Build detection content tuned to AI-generated malware artifacts (placeholder seed values, excessive defensive coding patterns, embedded emoji/unicode in binaries)
  • Expand threat-hunting coverage for MENA/Gulf critical infrastructure, aviation, and defense/aerospace/telecom sectors given active Iranian APT targeting
  • Coordinate with NGO/civil-society CERTs to distribute IOCs and takedown requests for shock-lure phishing infrastructure targeting activists

Affected and fixed versions

  • Various — Microsoft Office (macro-enabled documents)
    Vulnerable versions: all versions with macros enabled
    Fixed in: macros disabled by default / Protected View enforced
  • Various — Industrial Control Systems / PLCs (industrial refrigeration, energy, transportation, maritime)
    Vulnerable versions: internet-exposed ICS/OT devices
    Fixed in: network-segmented, non-internet-exposed configurations
  • Various — Aviation, defense, aerospace, and telecom sector enterprise networks (US, Europe, Middle East)
    Vulnerable versions: endpoints without EDR behavioral detection
    Fixed in: N/A - social-engineering/malware campaign, not a software vulnerability

Timeline

  • Iran begins an 88-day nationwide internet blackout, coinciding with expanded AI-assisted domestic surveillance and protest-crackdown infrastructure.
  • HarfangLab obtains RedKitten campaign malicious samples targeting Iranian human-rights NGOs and protest documentarians.
  • MuddyWater (GreenGolf) launches Operation Olalampo, targeting MENA organizations with phishing documents delivering GhostFetch, GhostBackDoor, HTTP_VIP, and the CHAR Rust backdoor.
  • HarfangLab publishes analysis of the RedKitten campaign and its AI-accelerated tooling (SloppyMIO implant).
  • US and Israeli forces conduct coordinated airstrikes on Iran (Operation Epic Fury / Roaring Lion), triggering surge in Iranian state and hacktivist cyber activity.
  • Dust Specter (APT34/OilRig-linked) targets Iraqi government officials using compromised Iraqi government infrastructure, deploying SPLITDROP/TWINTASK/TWINTALK and the consolidated GHOSTFORM RAT.
  • Nimbus Manticore resurfaces during Operation Epic Fury targeting aviation, defense, aerospace, and telecom organizations with the AI-assisted MiniFast backdoor.
  • Iran claims a strike on the USS Abraham Lincoln; an explosive drone boat targets vessel MKD VYOM.
  • Alleged Iranian Arash-2 strike on Ben Gurion Airport.
  • IRGC Navy Commander Tangsiri killed in an Israeli strike.
  • ChatGPT-refined attack reported against the Vyncs GPS tracking service.
  • Nimbus Manticore adds SEO poisoning (fake SQL Developer installer pages) as a MiniFast delivery vector.
  • Recorded Future publishes 'Iran's AI-Enhanced Asymmetric Playbook,' synthesizing the January-June 2026 campaign.
  • The US-Iran memorandum of understanding collapses, marking the close of the analyzed conflict period.

Sources cited

Showing 15 of 18 sources, primary sources first.

More from APT42

Further reading

Drafted by the Threadlinqs research pipeline from 18 sources; published ; last updated —.

Threadlinqs Intelligence — Real-Time Threat Detection Platform

// threat_feed
$ sort --newest
Showing all threats

Live intelligence console

Threat level

Latest Threats