Zscaler ThreatLabz 2026 Report: Encrypted Phishing & AiTM/BiTM Initial-Access Campaigns Targeting the Public Sector

Zscaler ThreatLabz 2026 Report (TL-2026-0878), also tracked as One Click to Compromise, is a high-severity phishing campaign, first published 2026-06-19. It has no confirmed attribution, affects Microsoft Microsoft 365 / Entra ID identity (most-impersonated brand, maps to 18 MITRE ATT&CK techniques (T1056.003, T1078, T1078.004), and is covered by 9 detection rules and 18 indicators of compromise.

Key facts for TL-2026-0878

Threat ID
TL-2026-0878
Also known as
One Click to Compromise, ThreatLabz 2026 Phishing and Initial Access Report
Severity
HIGH
Status
ACTIVE
Category
PHISHING
First published
2026-06-19
Last reviewed
2026-06-19
Attribution confidence
NONE
Motivation
FINANCIAL
Target sectors
government, healthcare, education, financial, services
Target regions
North America, Europe, South America
Detection rules
9
Indicators of compromise
18

Malware and tooling in Zscaler ThreatLabz 2026 Report

Malware and tooling: AiTM reverse-proxy / BiTM remote-browser relay, BlackForce, EvilProxy, Evilginx, Sneaky 2FA, Tycoon 2FA

Zscaler ThreatLabz's 2026 Phishing and Initial Access Report documents a 50% year-over-year surge in phishing against government (138.5M hits in 2025, the third-most-targeted industry), with 95.2% of phishing now delivered over encrypted channels. Adversary-in-the-middle (AiTM) and browser-in-the-middle (BiTM) phishing-as-a-service kits capture credentials and live MFA session tokens, while AI website-builder tools mass-produce high-fidelity phishing infrastructure.

How Zscaler ThreatLabz 2026 Report works

The Zscaler ThreatLabz 2026 Phishing and Initial Access Report ("One Click to Compromise") and its public-sector companion analysis describe a structural shift in initial-access tradecraft: overall phishing volume in the Zscaler cloud fell ~20% year over year (in both 2024 and 2025), but the surviving campaigns are more targeted, more AI-assisted, and harder to distinguish from legitimate activity. Government phishing rose 50% YoY to 138.5 million hits in 2025, ranking government the third-most-targeted industry; the Services sector saw the largest surge, climbing 65.5% from 330.9M to 547.7M hits. 95.2% of all phishing activity is now delivered over encrypted (TLS) channels, defeating defenses that do not perform consistent encrypted-traffic inspection.

The report's central tradecraft finding is the operationalization of adversary-in-the-middle (AiTM) and browser-in-the-middle (BiTM) phishing kits. These kits position a reverse proxy (or a remotely controlled browser session) between the victim and the legitimate identity provider, relaying the authentication flow in real time to capture both credentials and the post-MFA session cookie. Because the live session token is stolen after the MFA challenge succeeds, the attacker turns a single click into session-level compromise and bypasses legacy MFA (SMS one-time codes, TOTP, and push notifications). ThreatLabz cites BlackForce as an example phishing-as-a-service platform in this ecosystem; corroborating public research documents the broader AiTM PhaaS market — Tycoon 2FA (attributed to Storm-1747), EvilProxy, Sneaky 2FA, NakedPages, and the open-source Evilginx framework — sold for roughly $100-$1,000 per month and mapping to MITRE T1557 (Adversary-in-the-Middle), T1539 (Steal Web Session Cookie), and T1556.006 (Modify Authentication Process: MFA).

A second major theme is AI-generated phishing infrastructure. ThreatLabz identified 413,524 AI-generated site instances across the analysis period, of which 37,447 (9.06%) were flagged as malicious. The generation platforms most attributed were Manus AI (15.6%), Blackbox AI (14.3%), and Anything AI (9.8%), alongside DeepSite AI and Lovable AI. These tools let adversaries spin up pixel-accurate brand-impersonation sites in minutes rather than days. A specific observed campaign impersonated Brazilian government services (paired with instant-payment-system credential theft) using DeepSite AI and Blackbox AI. Microsoft and Google remain the most-impersonated brands, abused via credential-harvesting pages, SEO poisoning, and QR-code redirects.

The report also frames reconnaissance as a leading indicator of intrusion: ThreatLabz recorded 89.9 million hostile interactions with external decoys over six months and 121,000+ distinct AWS-hosted IPs probing customer environments. For the public sector specifically, healthcare saw 1.58 million encrypted attack hits while education recorded 1.6 billion encrypted attack hits over the year despite a 65.6% YoY decline in phishing attempts. Recommended defenses center on consistent TLS inspection, FIDO2/phishing-resistant authentication, identity-based least-privilege access that minimizes application exposure, monitoring for AI-generated phishing infrastructure, and extending visibility beyond the inbox to browsers, authentication flows, and SaaS sessions.

MITRE ATT&CK techniques used in TL-2026-0878

Credential Access

T1056.003 Web Portal Capture; T1539 Steal Web Session Cookie; T1556.006 Multi-Factor Authentication; T1557 Adversary-in-the-Middle

Persistence

T1078 Valid Accounts

Initial Access

T1078.004 Cloud Accounts; T1189 Drive-by Compromise; T1566 Phishing; T1566.002 Spearphishing Link

command-and-control

T1102 Web Service; T1573 Encrypted Channel

Command and Control

T1102.002 Bidirectional Communication

Resource Development

T1583 Acquire Infrastructure; T1588.002 Tool; T1608.005 Link Target

Reconnaissance

T1589 Gather Victim Identity Information; T1595 Active Scanning

stealth

T1684.001 Impersonation

Affected products and versions in Zscaler ThreatLabz 2026 Report

  • Microsoft — Microsoft 365 / Entra ID identity (most-impersonated brand; AiTM session-cookie theft target)
    Vulnerable versions: legacy MFA: SMS OTP, TOTP, push notifications
    Fixed in: FIDO2 / WebAuthn phishing-resistant authentication
  • Google — Google Workspace identity (most-impersonated brand; AiTM session-cookie theft target)
    Vulnerable versions: legacy MFA: SMS OTP, TOTP, push notifications
    Fixed in: FIDO2 / WebAuthn phishing-resistant authentication
  • Public Sector — Government, healthcare, and education organizations (third-most-targeted industry; Brazilian government services impersonation observed)
    Vulnerable versions: organizations without consistent TLS inspection or phishing-resistant MFA
    Fixed in: Zero Trust + FIDO2 + encrypted-traffic inspection

Remediation for Zscaler ThreatLabz 2026 Report

Immediate actions

  • Inspect encrypted/TLS traffic consistently at the proxy/gateway so the 95.2% of phishing delivered over encrypted channels is not invisible
  • Hunt for and revoke anomalous active session cookies; enforce short session lifetimes and re-authentication on sensitive actions to blunt AiTM session-token replay
  • Block known AiTM/BiTM PhaaS and AI-website-builder infrastructure (Manus AI, Blackbox AI, Anything AI, DeepSite AI, Lovable AI generated sites flagged malicious)

Workarounds

  • Conditional-access policies binding sessions to device compliance and network/location to reduce stolen-cookie replay value
  • Block or sandbox QR-code links and newly-registered look-alike domains; restrict use of unmanaged browsers for privileged identity flows

Longer-term hardening

  • Deploy FIDO2/WebAuthn phishing-resistant authentication to remove the most reliable click-to-session-compromise path
  • Apply identity-based least-privilege (Zero Trust) access controls that minimize public application exposure
  • Extend monitoring beyond the inbox to browsers, authentication flows, and SaaS sessions; deploy deception/decoy technology for early reconnaissance detection

Weaknesses (CWE) in Zscaler ThreatLabz 2026 Report

CWE-1021, CWE-290, CWE-294, CWE-384, CWE-451

Timeline of Zscaler ThreatLabz 2026 Report

  • Tycoon 2FA AiTM phishing-as-a-service kit first observed (attributed to Storm-1747), industrializing real-time MFA-bypass tradecraft later seen across the 2026 ecosystem.
  • Overall phishing volume in the Zscaler cloud begins its ~20% year-over-year decline, continuing through 2024 and 2025 as campaigns shift to targeted, AI-assisted operations.
  • 2025 telemetry period analyzed: government phishing reaches 138.5 million hits (+50% YoY); Services industry climbs 65.5% from 330.9M to 547.7M hits.
  • Six-month window recording 89.9 million hostile interactions with external decoys and 121,000+ distinct AWS-hosted IPs probing customer environments.
  • 413,524 AI-generated site instances identified; 37,447 (9.06%) flagged malicious, attributed to Manus AI (15.6%), Blackbox AI (14.3%), Anything AI (9.8%), DeepSite AI and Lovable AI.
  • Observed campaign impersonating Brazilian government services using DeepSite AI and Blackbox AI, paired with instant-payment-system credential theft.
  • Microsoft publishes in-depth analysis of Tycoon2FA AiTM operations at scale, corroborating the PhaaS MFA-bypass tradecraft described by ThreatLabz.
  • Zscaler ThreatLabz publishes the 2026 Phishing and Initial Access Report ('One Click to Compromise') and press release noting a nearly 50% increase in phishing with education, finance, and government most targeted.
  • Zscaler publishes the public-sector companion analysis: 95.2% of phishing over encrypted channels; government third-most-targeted; healthcare 1.58M and education 1.6B encrypted attack hits.

Sources cited for Zscaler ThreatLabz 2026 Report

Threats related to Zscaler ThreatLabz 2026 Report

Detection coverage for TL-2026-0878

As of 2026-06-19, Threadlinqs Intelligence publishes 9 detection rule(s) for TL-2026-0878 across Splunk SPL, Microsoft KQL and Sigma, covering 18 indicator(s) of compromise. The whole corpus is readable without an account; a free account unlocks full detection query text in Splunk SPL, Microsoft KQL and Sigma; paid tiers add raw indicator values, correlation and the MCP server. Threadlinqs MCP server · View plans.

Further reading

Threadlinqs Intelligence — Real-Time Threat Detection Platform

[ 0 threats ] [ 0 det ] [ CRIT: 0 ] [ HIGH: 0 ]
// threat_feed
$ sort --newest
Showing all threats

Latest Threats