KENSAI Security Research

Critical deserialization vulnerability in Windchill and FlexPLM product lifecycle management systems allowing unauthenticated remote code execution. The severity is unprecedented — German federal police (BKA) dispatched agents to companies nationwide over the weekend, waking system administrators at night to hand-deliver the advisory. PTC has published IoCs including webshell indicators (GW.class, dpr_*.jsp files) and suspicious user-agent patterns. No patch available yet; vendor recommends applying Apache/IIS servlet path denial rules immediately and disconnecting internet-facing instances.
CRITICALZERO-DAYRCEPLMMANUFACTURING
→ BleepingComputer

An out-of-bounds read vulnerability in Citrix NetScaler can be exploited remotely without authentication to read sensitive information from memory. Multiple security firms warn exploitation is imminent given the widespread deployment of NetScaler appliances. Organizations should patch immediately.
CRITICALPRE-AUTHNETSCALERNETWORK
→ SecurityWeek

Google released Chrome 146 patching eight memory safety bugs affecting seven different components. Organizations should push browser updates immediately across fleets.
HIGHBROWSERMEMORY-SAFETY
→ SecurityWeek

The TeamPCP threat actor — behind the recent Trivy and KICS compromises — has now backdoored the popular LiteLLM Python package on PyPI. The payload is a three-stage attack:

Stage 1: Credential harvester sweeping SSH keys, cloud credentials, K8s secrets, crypto wallets, and .env files
Stage 2: Kubernetes lateral movement toolkit deploying privileged pods to every node
Stage 3: Persistent systemd backdoor (sysmon.service) polling a C2 domain for additional binaries

Malicious versions have been removed from PyPI. Anyone who installed litellm 1.82.7 or 1.82.8 should consider their environment fully compromised.
CRITICALSUPPLY-CHAINPYPIK8SCREDENTIAL-THEFT
→ The Hacker News

TeamPCP has compromised two Checkmarx-maintained GitHub Actions workflows (checkmarx/ast-github-action and checkmarx/kics-github-action) using stolen CI credentials — likely obtained from the earlier Trivy supply chain breach. This is the same threat actor expanding their reach through interconnected CI/CD dependencies.
CRITICALSUPPLY-CHAINCI/CDGITHUB-ACTIONS
→ The Hacker News

Seven npm packages published by user "mikilanjillo" are stealing cryptocurrency wallets and credentials. Packages include names like react-performance-suite, react-state-optimizer-core, react-fast-utilsa, and ai-fast-auto-trader — designed to look like legitimate React and AI utility libraries.
HIGHSUPPLY-CHAINNPMCRYPTO-THEFT
→ The Hacker News

Hackers stole personal information, medical records, and health insurance data from QualDerm's internal systems. 3.1 million individuals affected — one of the larger healthcare breaches this quarter.
BREACHHEALTHCARE3.1M RECORDS
→ SecurityWeek

Bug bounty platform HackerOne is notifying hundreds of employees that their data was stolen after attackers compromised Navia, a U.S. benefits administrator used by the company. This is a third-party/supply chain data breach — the attacker didn't hit HackerOne directly.
BREACHTHIRD-PARTYHR-DATA
→ BleepingComputer

Widely used K-12 student information system Infinite Campus is warning customers of a data breach following an extortion attempt by the ShinyHunters threat actor. Student and school data at risk.
BREACHEDUCATIONEXTORTION
→ BleepingComputer

The Lapsus$ extortion group claims to have breached AstraZeneca, allegedly compromising internal code repositories, credentials, and employee data. Under investigation.
BREACHPHARMAEXTORTION
→ SecurityWeek

The Dutch Ministry of Finance disclosed a breach affecting employee data. Details still emerging, but this adds to a pattern of European government entities being targeted.
BREACHGOVERNMENTEU
→ BleepingComputer

Hackers stole internal IDs, names, email addresses, and business partner IDs from an internal management system. Impact scope still being assessed.
BREACHAUTOMOTIVE
→ SecurityWeek

A Russian national was sentenced to nearly 7 years (81 months) in U.S. federal prison after pleading guilty to acting as an initial access broker (IAB) for the Yanluowang ransomware group. The operations caused approximately $9M in damages. A win for international law enforcement cooperation.
RANSOMWARELAW-ENFORCEMENTSENTENCING
→ BleepingComputer

An ongoing phishing campaign targets French-speaking corporate environments using obfuscated VBScript files disguised as résumés. The multi-purpose toolkit combines credential theft, data exfiltration, and Monero mining. Uses Dropbox for staging, Moroccan WordPress sites for C2, and mail.ru SMTP for exfil.
HIGHPHISHINGCRYPTOMININGINFOSTEALER
→ The Hacker News

A large-scale malvertising campaign abuses Google Ads to target US users searching for tax documents. Delivers rogue ConnectWise ScreenConnect installers that deploy "HwAudKiller" — a tool that uses Bring Your Own Vulnerable Driver (BYOVD) technique with a Huawei driver to disable EDR/AV products. Active since January, likely to intensify as US tax deadline approaches.
HIGHMALVERTISINGBYOVDEDR-BYPASS
→ The Hacker News

The Federal Communications Commission has updated its Covered List to include all consumer routers made in foreign countries, banning the sale of new models in the United States. A significant escalation in the US approach to hardware supply chain security.
POLICYSUPPLY-CHAINHARDWARE
→ BleepingComputer

Mozilla released Firefox 149 with a built-in VPN offering up to 50GB of monthly traffic at no cost. A notable move toward privacy-by-default in mainstream browsers.
TOOLPRIVACYBROWSER
→ BleepingComputer

RSA Conference 2026 has kicked off with a wave of vendor announcements. Key themes emerging: Guardian Agents (AI agents that supervise other AI agents — Gartner published its first Market Guide), agentic AI governance, and continued convergence of identity + cloud security.
INDUSTRYRSACAI-SECURITY
→ SecurityWeek

Gartner released its inaugural Market Guide for Guardian Agents — AI agents that supervise other AI agents to ensure their actions align with goals and boundaries. This new category is directly relevant as agentic AI platforms shift from passive tools to autonomous action-takers with real system access.
MARKETAI-GOVERNANCEGARTNER
→ The Hacker News

XM Cyber's threat research team mapped eight attack vectors inside AWS Bedrock environments. When AI agents can query Salesforce, trigger Lambda functions, or pull from SharePoint knowledge bases, they become nodes in infrastructure with permissions and reachable paths to critical assets. Research highlights the expanding attack surface of agentic AI in cloud environments.
HIGHCLOUDAI-SECURITYAWS
→ The Hacker News

Poland reports a major surge in cyberattacks throughout 2025, including a destructive infiltration of the country's energy system in December, suspected to originate from Russia. The retrospective report underscores ongoing state-sponsored threats to European critical infrastructure.
STATE-SPONSOREDENERGYEU
→ SecurityWeek

Israel reportedly hijacked Iran's national street camera surveillance network, originally built to suppress dissent, and repurposed it as a targeting tool. The role in the killing of Iran's supreme leader underscores how surveillance systems are becoming warfare assets.
CYBER-WARFARESURVEILLANCESTATE-SPONSORED
→ SecurityWeek

The FBI published an alert describing malware used by Iranian government hackers in an attack on Stryker. A malicious file was found during the investigation. Details of the Iran-linked tooling are now public for threat intel purposes.
STATE-SPONSOREDIRANMALWARE
→ SecurityWeek

1. Supply Chain Attacks Are Accelerating
TeamPCP's cascading compromise (Trivy → KICS → LiteLLM → Checkmarx) shows how a single CI/CD breach can propagate through interconnected open-source tooling. The "Ghost" npm campaign adds to the pattern. Expect more.

2. AI Tooling Is the New Attack Surface
LiteLLM (AI proxy), AWS Bedrock attack vectors, Guardian Agent governance — three separate stories all point to agentic AI infrastructure becoming a prime target. Organizations deploying AI agents need security oversight now.

3. BYOVD Is Going Mainstream
The tax season malvertising campaign using a Huawei driver to disable EDR shows BYOVD (Bring Your Own Vulnerable Driver) is no longer an APT-only technique — commodity malware campaigns are adopting it.

4. European Governments Under Sustained Pressure
Dutch Ministry of Finance breach, Polish energy sector attacks, and the PTC Windchill emergency (which triggered BKA mobilization) paint a picture of EU infrastructure under persistent threat — reinforcing the NIS2 compliance urgency.

5. Third-Party/Vendor Risk Remains Underestimated
HackerOne breached through a benefits administrator. AstraZeneca allegedly breached. Mazda through an internal management system. The attack path through vendors and partners continues to be the weakest link.