← Back to Briefings
DAILY BRIEFING · JUNE 17, 2026 · #091

Fortinet actively exploited; Rokarolla targets 217 banking apps; Google Vertex AI flaw

📅 June 17, 2026🤖 AI-Generated Analysis5 min read
Severity High
How to read this briefing
Verified facts — NVD & CISA KEV Partially verified — awaiting NVD enrichment AI analysis — synthesis, verify before acting [1]Inline citations — click any [N] to view the source
How our verification pipeline works →
Actionable · Verified facts
NVD-published · CISA KEV cross-checked
CVECVSSVendor · ProductExploitationRefs
🛡️CVE-2026-398139.8 NVD 3.1Fortinet Fortisandbox In the wild[1]
🛡️CVE-2026-398089.8 NVD 3.1Fortinet Fortisandbox In the wild[1] [2]
🛡️CVE-2026-250899.8 NVD 3.1Fortinet FortiSandbox In the wild[1]
Contextual · AI analysis Synthesized from 10 feeds · verify before acting

TL;DR

Fortinet FortiSandbox actively exploited; three CVEs confirmed in the wild. Android malware Rokarolla targets 217 banking/crypto apps with 137 commands. Google Vertex AI SDK bucket-squatting flaw allows unauthorized model hijacking and code execution.

THREAT LEVEL: HIGH – Multiple actively exploited vulnerabilities and emerging mobile banking trojans require immediate patching and monitoring.

Executive Summary

Top Threats Today

1. Fortinet FortiSandbox Active Exploitation

Severity: HIGH   Affected: Technology

Threat actors are actively exploiting three security vulnerabilities in Fortinet FortiSandbox, according to threat intelligence firm Defused Cyber [1]. The exploited CVEs are CVE-2026-39813, CVE-2026-39808, and CVE-2026-25089, with exploitation activity observed over the past 24 hours [1]. One of the three vulnerabilities was patched within the past week [1].
Sources:[1] The Hacker News

Recommended Action

  • Immediately verify FortiSandbox version and apply all available security patches
  • Review FortiSandbox logs for indicators of exploitation (unusual network connections, failed authentication attempts, unexpected process execution)
  • Isolate affected FortiSandbox instances from network if patching cannot be applied immediately
  • Monitor threat intelligence feeds for detailed exploitation signatures and IOCs

2. Rokarolla Android Banking Trojan Deployment

Severity: HIGH   Affected: Finance

Security researchers at Zimperium's zLabs have documented a new Android banking trojan named Rokarolla that targets 217 banking and cryptocurrency applications [1]. The malware packs 137 remote commands, providing operators near-total control of infected devices [1]. The trojan is capable of lifting lock-screen PINs and reading SMS codes used in two-factor authentication [1]. Rokarolla has been spread via fake TikTok and Chrome downloads .
Sources:[1] The Hacker News[2] BleepingComputer

Recommended Action

  • Deploy mobile device management (MDM) policies blocking sideloading and restricting app installation to official app stores
  • Educate users to download banking and financial applications exclusively from official app stores (Google an unattributed threat actor, Apple App Store)
  • Enable SMS-based OTP notifications on accounts and monitor for suspicious SMS activity
  • Implement application allow-listing on managed Android devices for known legitimate banking apps

3. Google Vertex AI SDK Bucket Squatting Vulnerability

Severity: HIGH   Affected: Technology

A flaw in the Google Cloud Vertex AI SDK for Python allows an attacker with no access to a victim's project to hijack the victim's machine learning model upload and execute code inside Google's serving infrastructure [1]. The vulnerability leverages bucket squatting techniques to intercept model uploads [1]. Palo Alto Networks Unit 42 discovered and reported the bug through Google's bug bounty program [1].
Sources:[1] The Hacker News

Recommended Action

  • Update Google Cloud Vertex AI SDK to the patched version immediately
  • Audit Google Cloud Storage bucket naming conventions and apply restrictive IAM policies
  • Enable bucket versioning and object hold policies to prevent unauthorized overwrites
  • Monitor Cloud Storage audit logs for suspicious upload and access patterns

4. ClickFix Campaign Expansion with New Loaders

Severity: HIGH   Affected: Technology

Cybersecurity researchers have identified multiple ClickFix campaigns delivering three new malware loaders: BabaDeda Loader, Lorem Ipsum Loader, and Potemkin [1]. Attacks using BabaDeda Loader were observed in April 2026 [1]. The Lorem Ipsum malware variant pivots to ClickFix delivery via compromised WordPress sites and may be linked to the ransomware and data extortion group Vice Society [2].
Sources:[1] The Hacker News[2] Dark Reading

Recommended Action

  • Deploy URL filtering to block known ClickFix distribution domains and fake update sites
  • Implement application whitelisting to restrict execution of unsigned or untrusted binaries
  • Train users to verify software updates through official vendor channels and never download updates via unsolicited prompts
  • Monitor for browser extensions and plugins requesting unusual permissions

5. Malicious JetBrains Marketplace Plugins Stealing API Keys

Severity: HIGH   Affected: Technology

At least 15 malicious plugins found on the JetBrains Marketplace were designed to steal AI API keys from developers [1]. The plugins target development environments where sensitive credentials are stored and used for API authentication.
Sources:[1] BleepingComputer

Recommended Action

  • Audit all installed JetBrains IDE plugins and remove any plugins from untrusted or unknown publishers
  • Rotate all AI API keys and credentials that may have been exposed through installed plugins
  • Implement IDE plugin governance policies restricting plugin installation to approved vendors only
  • Enable API key rotation alerts and monitor for unauthorized API usage patterns

Today’s Action Checklist

🤖 This briefing was compiled by defend.network using AI-powered analysis of multiple cybersecurity sources including CISA advisories, vendor security bulletins, and threat intelligence feeds. Always verify critical intelligence through official vendor channels before taking action.

Get Tomorrow’s Briefing in Your Inbox

Subscribe free and never miss a daily threat briefing.