Lyrie
Defensive-Playbook
0 sources verified·11 min read
By Lyrie Threat Intelligence·4/27/2026

Before the Ransom Demand: The Definitive 2026 Defensive Playbook Against Initial Access Brokers and the Infostealer Pipeline

TL;DR

  • Cisco Talos Q1 2026 IR data: phishing is back as the #1 initial access vector (>33% of engagements), reversing Q3-Q4 2025 exploit-dominated trends.
  • Flare 2026 infostealer report: 16% of infostealer-infected endpoints expose enterprise SSO credentials — up from 6% in 2024. 79% of logs contain Microsoft Entra ID credentials. 1.17 million logs carry credentials plus session cookies, enabling MFA bypass.
  • IABs price access from $20 (low-privilege credentials) to $10,000+ (domain admin with persistence) on dark web forums. This is the invisible toll booth between phishing and ransomware.
  • The Q1 2026 silver lining: pre-ransomware incidents dropped to 18% of Talos engagements (from 50% in Q1 2025) — because defenders caught it early. This playbook tells you how.
  • Lyrie's autonomous detection layer catches IAB-stage activity at machine speed — before the ransomware operator ever buys the access.

Background: The Assembly Line You're Not Defending Against

There is a criminal supply chain operating in plain sight, and most enterprise security programs are optimized for the wrong end of it.

The conventional mental model of ransomware looks like this: attacker phishes employee → installs ransomware → demands payment. The reality in 2026 is a three-tier assembly line with distinct specialists at each stage, and the critical handoff happens in a marketplace that 98% of organizations have no visibility into.

Tier 1 — Initial Access Brokers (IABs): Specialists in breaking in and nothing else. They gain unauthorized access to corporate networks — via unpatched VPNs, brute-forced RDP, phishing, infostealer deployments, or insider bribery — then package and sell that access on underground forums. They take no further action on the target. Their business model is volume, speed, and reputation.

Tier 2 — Access Validators: Mid-tier actors who buy cheap credential dumps and add value by validating which credentials are active, escalating privileges, and repackaging as premium listings. A $20 bulk credential set becomes a $500 validated enterprise account entry.

Tier 3 — Ransomware Operators / RaaS Affiliates: The buyers. They purchase confirmed access, deploy their toolkit (SystemBC for persistence, Cobalt Strike for lateral movement, custom ransomware for encryption), and execute. Their timeline from access purchase to ransomware deployment has compressed to under 48 hours in 2026.

Understanding this pipeline is prerequisite to disrupting it. You cannot defend against ransomware if you're only looking for ransomware.


The 2026 Infostealer Explosion: How Your Credentials Became a Commodity

Infostealers are the primary feeding mechanism for the IAB economy. These lightweight malware variants — RedLine, Lumma, Rhadamanthys, Vidar, and their dozens of variants — silently harvest browser-saved passwords, session cookies, authentication tokens, and cloud credentials from infected endpoints. They require no persistence, no C2 infrastructure beyond initial upload, and no complex deployment. A credential-stealer infection can be live and exfiltrating within seconds.

The scale is industrial. Flare's 2026 threat landscape report documents:

  • 1.17 million infostealer logs sold on dark web markets containing both credentials AND active session cookies — enabling attackers to bypass MFA entirely by replaying authenticated sessions.
  • 79% of logs contain Microsoft Entra ID credentials — meaning the average infostealer infection is now effectively an Azure/M365 breach.
  • 16% of infostealer-infected endpoints expose enterprise SSO credentials — triple the 2024 rate of 6%. The jump correlates directly with the explosion of corporate BYOD endpoints running personal browsers with corporate credentials saved.
  • Logs are priced as low as $5-10 per infection on bulk markets, with enterprise SSO credentials fetching $100-500 per log depending on organization size and privilege level.

The geometry of this is alarming: a threat actor spending $500 on infostealer logs is buying a probability distribution across hundreds of corporate environments. They don't target — they spray and sort.

What the attacker does next depends entirely on what the log contains:

  • Low-privilege credentials → sold cheap to other IABs or used for BEC fraud
  • Corporate VPN credentials → priced premium, often sold exclusive to ransomware operators
  • Session cookies for cloud consoles → used immediately for cloud resource hijacking or data exfiltration
  • Domain admin credentials → top-tier listing, often auctioned, prices in four-to-five figures

The IAB Marketplace: How Access Gets Priced and Sold

IABs operate on dark web forums (RAMP, BreachForums successor sites, Russian-language underground markets) with reputation systems, escrow services, and even customer support. This is not a chaotic bazaar — it is a functioning market with consistent pricing dynamics.

Pricing tiers as documented by Orange Cyberdefense and KnowBe4 research (April 2026):

| Access Type | Price Range | Common Source |

|---|---|---|

| Low-privilege user credentials (unvalidated) | $5–$20 | Bulk infostealer dumps |

| Validated corporate email/VPN credentials | $30–$500 | Validated infostealer logs |

| RDP access (low privilege) | $50–$200 | Brute force, exposed RDP |

| Network foothold with persistence | $500–$3,000 | Post-exploitation IAB |

| Domain admin / full network control | $5,000–$50,000+ | Premium IAB with privilege escalation |

| Critical infrastructure / healthcare / government | $10,000–$100,000 | Exclusive, auctioned |

The critical insight: IABs sell speed. Their listings go stale. A VPN credential is only valuable until the password is rotated or the patch is applied. This creates urgency for buyers — and a detection window for defenders. Between the time an IAB posts access for sale and when a ransomware operator deploys their toolkit, there is typically a 12-72 hour window.

That window is your hunting opportunity.


Technical Analysis: The IAB Attack Surface in 2026

Primary Entry Vectors

1. Phishing with AI-Generated Lures

Cisco Talos Q1 2026 confirmed the first documented use of an AI tool (Softr) in a phishing campaign targeting Microsoft Exchange and OWA credentials in the public administration sector. Using AI-powered "vibe coding," threat actors generated convincing credential harvesting pages in minutes with no code. The data was automatically routed to Google Sheets — a disposable exfiltration channel with no malicious infrastructure to detect.

This represents the democratization of sophisticated phishing: what previously required HTML/CSS skill and hosting infrastructure now requires a free Softr account and three prompts. Expect volume to scale dramatically.

2. Unpatched Internet-Facing Systems

Every unpatched VPN appliance, RDP endpoint, Exchange server, and web application is a potential IAB product listing. In Q1 2026, IABs actively tracked:

  • Ivanti Connect Secure remnant vulnerabilities
  • Fortinet FortiGate authentication bypass chains
  • Cisco ASA / RVS command injection flaws
  • SonicWall SSL-VPN session fixation bugs
  • Microsoft Exchange ProxyLogon/ProxyShell derivative chains still active in unpatched environments

The IAB workflow is semi-automated: scanning tools (Shodan, Censys, FOFA) identify exposed services; exploit frameworks (Metasploit, custom tooling) are applied at scale; working sessions are captured and packaged.

3. MFA Fatigue and Session Cookie Replay

With 1.17 million infostealer logs containing active session cookies, MFA has effectively been bypassed at scale before the attacker even engages. Cookie replay attacks against Microsoft 365, Okta, and Google Workspace don't trigger MFA prompts — the attacker presents a valid authenticated session. Detection requires behavioral analysis, not authentication monitoring.

MFA fatigue attacks (Cisco, Uber, MGM breach pattern) remain effective against organizations using push-based MFA without number matching.

The New Actor: Crimson Collective

Talos Q1 2026 documented a first encounter with Crimson Collective, a newly identified threat group. Specific TTPs were not fully disclosed in the public summary, but the group's appearance in IR engagements targeting public administration — alongside the Softr phishing campaign — suggests an IAB-adjacent or affiliate-model operator entering the market with AI-assisted tooling. Watch for further attribution.


IOCs / Indicators

Behavioral indicators of IAB-stage activity (pre-ransomware):

  • Authentication from unexpected geographies on corporate SSO (Entra ID, Okta) — particularly Eastern Europe, Southeast Asia, and VPN/proxy exit nodes
  • Browser-saved credential exfiltration patterns: abnormal reads of %APPDATA%\Local\[Browser]\User Data\Default\Login Data
  • Anomalous network connections from endpoint processes to Telegram API, Discord webhooks, or paste-site domains (infostealer C2/exfil channels)
  • Unexpected Scheduled Task creation or WMI subscription events following a phishing click
  • Dark web mentions of your organization's name, domain, or ASN on IAB marketplaces (requires threat intel subscription)
  • RDP access from new endpoints not in baseline — especially outside business hours
  • Entra ID sign-in logs: successful authentications with tokenSatisfied=false (cookie replay indicator)

Infostealer malware family indicators:

  • Lumma Stealer: C2 domains cycling through .shop, .top, .xyz TLDs; PowerShell download cradles from mshta.exe
  • RedLine/RecordBreaker: Process injection into RegAsm.exe, MSBuild.exe; string patterns grabber|stealer|clipper in memory
  • Rhadamanthys: .NET loader in %TEMP%, connects to bulletproof hosting in AS9009 (M247)

Lyrie Take

The IAB layer is where machine-speed defense wins.

The gap Lyrie is built to close is precisely this: the time between an IAB listing appearing on a dark web forum and a ransomware operator purchasing it and deploying. Human analysts monitoring dark web forums can detect this — eventually. By the time a threat intelligence report arrives in your inbox, the operator already has admin access.

Lyrie's autonomous detection pipeline ingests dark web telemetry feeds, correlates against your environment's asset inventory and identity surface, and flags IAB-relevant signals in real time. When a forum listing appears claiming VPN access to an ASN matching your IP range — or when an infostealer log surfaces containing a credential pattern matching your domain — the alert fires in seconds, not hours.

More critically: Lyrie detects IAB-stage behavior on the endpoint before exfiltration completes. The infostealer's credential harvesting touches predictable file paths, makes predictable network calls, and generates predictable process trees. At machine speed, these patterns are unambiguous. The attacker's $20 investment never pays off.

This is the autonomous defense thesis: not reacting to ransomware, but intercepting the $20 transaction that funds it.


Defender Playbook

Immediate (0–48 hours)

1. Audit your infostealer exposure

  • Subscribe to an infostealer log monitoring service (Flare, Hudson Rock, Constella) and run your domains against existing leak databases now. Assume you have credentials in the wild.
  • Identify which employees have had personal devices infected with infostealers — cross-reference with corporate SSO login anomalies.

2. Harden VPN and RDP attack surface

  • Enumerate internet-facing RDP (Shodan query: port:3389 org:"[YOUR ASN]"). Any exposed RDP that isn't behind a VPN is an IAB product. Close it.
  • Audit VPN appliance patch levels. Anything Ivanti, Fortinet, Cisco, or SonicWall that isn't on the latest firmware is a potential listing.
  • Enable certificate-based authentication on VPN where possible; deprecate password-only auth.

3. Session cookie protection

  • Enable Continuous Access Evaluation (CAE) in Microsoft Entra ID — forces re-authentication on IP change, revoking replayed session cookies.
  • Audit conditional access policies for gaps: ensure "require compliant device" is enforced for all Exchange, SharePoint, and Azure portal access.
  • Enable token protection where supported (Entra ID Token Protection Preview).

Short-term (1–2 weeks)

4. Implement number-matching MFA

  • Disable push-notification-only MFA across all identity providers. Require number matching or FIDO2 hardware keys for privileged accounts.
  • For Okta: enable Okta FastPass with biometric verification as default.
  • For Entra ID: enforce number matching in Authentication Methods policy.

5. Deploy SIEM detection rules for IAB-stage behavior

Sigma rules to implement immediately:

# Infostealer credential database access
title: Browser Credential Store Access by Unusual Process
detection:
  selection:
    TargetFilename|contains:
      - '\Local\Google\Chrome\User Data\Default\Login Data'
      - '\Local\Microsoft\Edge\User Data\Default\Login Data'
      - '\Roaming\Mozilla\Firefox\Profiles\'
  filter:
    Image|endswith:
      - '\chrome.exe'
      - '\msedge.exe'
      - '\firefox.exe'
  condition: selection and not filter

# RDP access outside business hours from new source
title: RDP Logon Outside Business Hours - New Source IP
detection:
  selection:
    EventID: 4624
    LogonType: 10
  filter:
    SourceNetworkAddress|startswith:
      - '10.'
      - '192.168.'
  condition: selection and not filter and (hour > 20 or hour < 6)

6. Dark web monitoring

  • Configure monitoring for your organization's name, domain, key personnel names, and ASN range on IAB forum indexers.
  • Minimum: set up alerts on dehashed.com, IntelX, and Have I Been Pwned API for your domain.
  • Better: commercial threat intel (Flare, Recorded Future, Mandiant Advantage) with IAB marketplace coverage.

Structural (30 days)

7. Identity threat detection and response (ITDR)

  • Deploy an ITDR solution with behavioral baselines for every identity in your environment. Abnormal authentication patterns (new country, new device, new time-of-day) should auto-trigger step-up authentication or session revocation, not just generate alerts.
  • Vendors to evaluate: CrowdStrike Falcon Identity Protection, SentinelOne Singularity Identity, Microsoft Entra Identity Protection P2.

8. Privileged access hardening

  • Domain admin accounts must never have mail. Ever. Email is phishing surface.
  • Enforce Privileged Access Workstations (PAWs) for all tier-0 administration.
  • Implement LAPS (Local Administrator Password Solution) for all workstations — eliminates lateral movement via shared local admin passwords.

9. Supply chain credential hygiene

  • Audit SSO integrations. Every SaaS tool with Entra ID or Okta SSO is a potential pivot once the identity layer is compromised.
  • Revoke unnecessary OAuth grants. Review what apps have delegated access to your Microsoft 365 tenant (admin.microsoft.com → Integrated apps).
  • Enforce quarterly credential rotation for service accounts; implement secrets vaulting (HashiCorp Vault, Azure Key Vault).

10. Tabletop: Pre-ransomware response

  • Run a tabletop exercise specifically for the scenario: "We have been notified that VPN credentials for our organization are listed for sale on a dark web forum. It's 2 AM."
  • Decision points: forced password reset for all users vs. targeted reset? Take VPN offline? Engage forensics? Who has authority?
  • Talos Q1 2026 data shows that organizations that caught pre-ransomware indicators and acted did not deploy ransomware. The playbook works — when it exists.

Sources

1. Cisco Talos IR Trends Q1 2026 — "Phishing reemerges as top initial access vector" — blog.talosintelligence.com, April 22, 2026

2. Computing.co.uk — "Cybercrime: Who are the initial access brokers, and why do they sell us so cheap?" — April 2026 (sourcing KnowBe4, Orange Cyberdefense analysis)

3. Saptang Labs — "2026 External Threat Landscape: What CISOs Need to Know" — citing Flare 2026 infostealer report

4. Erdal Ozkaya — "Dark Web in 2026: What CISOs Need to Know" — erdalozkaya.com, April 2026

5. Adaptive Security — "Ransomware: Essential 2026 Guide for Cybersecurity Teams" — adaptivesecurity.com, April 2026

6. PurpleOps — "Ransomware Activity Tracker 2026" — purple-ops.io (IAB attribution patterns)

7. Microsoft — Continuous Access Evaluation documentation (CAE for session protection)

8. Sigma HQ — Sigma detection rule format reference


Lyrie Verdict: The IAB marketplace is the most underdefended stage in the ransomware kill chain. Your SIEM is tuned for ransomware deployment. Your SOC looks for encryption activity. The $20 credential sale that funded the entire operation happened in a window you never watched. Close the window first.

Lyrie.ai Cyber Research Division — Senior Analyst Desk

Lyrie Verdict

Lyrie's autonomous defense layer flags this class of exposure the moment it surfaces — no signature update required.