DFIR, deception, detection. Posts I wrote, intel my pipeline summarized, and redacted writeups from the fleet.
The article outlines the emerging security risks associated with autonomous Agentic AI and presents a collaborative architectural solution between TrendAI and NVIDIA. By integrating TrendAI's governance and behavioral analysis with NVIDIA's OpenShell runtime, enterprises can safely deploy self-evolving AI agents with runtime policy enforcement and protection against AI-native threats like prompt injection.
Unit 42 researchers demonstrated that both open and closed-source LLMs remain vulnerable to prompt jailbreaking at scale using a genetic algorithm-based fuzzing technique. By systematically generating meaning-preserving variants of disallowed requests, researchers successfully bypassed content filters and model guardrails, highlighting the fragility of current AI safety mechanisms under automated adversarial variation.
This report summarizes a LABScon 25 presentation detailing the sophisticated attack vectors used in cryptocurrency heists, which have resulted in $9 billion in losses. Threat actors are increasingly targeting developers and software supply chains—such as modifying production JavaScript code and compromising GitHub accounts via personal infrastructure—to execute massive wallet drains.
The Canadian Centre for Cyber Security released a daily digest highlighting critical vulnerabilities in Spring AI, including SQL and JSONPath injections, as well as unspecified vulnerabilities in GitHub Enterprise Server. Organizations utilizing these products are advised to apply the latest security patches to mitigate potential exploitation risks.
Boggy Serpens (MuddyWater) is conducting ongoing cyberespionage campaigns targeting critical infrastructure and diplomatic entities globally. The group leverages hijacked accounts for trusted relationship compromises, delivering advanced, AI-assisted malware toolkits including Rust-based backdoors and custom C2 protocols to maintain long-term persistence and evade detection.
The Warlock ransomware group (Water Manaul) has enhanced its attack chain by exploiting Microsoft SharePoint servers for initial access and deploying a sophisticated post-exploitation toolkit. The group leverages BYOVD techniques via the NSecKrnl.sys driver to disable security tools, establishes redundant C&C channels using legitimate tools like Velociraptor and Cloudflare Tunnels, and automates ransomware deployment domain-wide using Group Policy Objects (GPO).
Modern AI factories utilize massive, interconnected GPU clusters that generate high volumes of east-west traffic, rendering traditional perimeter and host-based security ineffective. To secure these environments without degrading performance, organizations must adopt infrastructure-level, identity-based microsegmentation using technologies like DPUs to enforce Zero Trust and contain lateral movement.
In 2025, ransomware operators increasingly relied on vulnerability exploitation for initial access and heavily targeted virtualization infrastructure like ESXi. While overall ransomware profitability appears to be declining, threat actors have adapted by increasing data theft extortion, targeting smaller organizations, and utilizing cross-platform ransomware families like REDBIKE, AGENDA, and INC.
A novel phishing campaign is abusing the legitimate LiveChat SaaS platform to impersonate brands like PayPal and Amazon. By engaging victims in real-time chat interfaces using automated bots or human operators, attackers successfully harvest sensitive information, including account credentials, multi-factor authentication (MFA) codes, personally identifiable information (PII), and credit card details.
Elastic has introduced open-source Agent Skills that enable AI coding agents to natively interact with Elastic Security. These skills allow security teams to rapidly provision cloud environments, generate realistic sample attack data, and manage alerts and detection rules directly from their IDEs.
CISA has added CVE-2025-47813, an information disclosure vulnerability in Wing FTP Server, to its Known Exploited Vulnerabilities (KEV) Catalog due to evidence of active exploitation. All organizations are strongly urged to prioritize timely remediation of this vulnerability to reduce exposure to cyberattacks.
The GlassWorm threat actor has evolved its supply chain attack methodology by abusing VS Code extension manifest fields to transitively deliver malicious payloads. This technique allows initially benign extensions to pull in malicious dependencies during later updates, executing staged JavaScript loaders that target developer workstations for credential and secret theft.
The 2025 Identity Threat Landscape Report highlights a massive surge in credential theft driven by infostealer malware, with LummaC2 leading the ecosystem. A critical finding is the widespread theft of active session cookies, which allows attackers to bypass multi-factor authentication (MFA) and directly access high-value corporate systems, VPNs, and cloud platforms.
Following the outbreak of a geopolitical conflict in the Middle East in early 2026, Akamai observed a 245% surge in malicious cyber activity targeting global enterprises. The threat landscape is characterized by massive increases in automated reconnaissance, credential harvesting, and data-wiping attacks by state-sponsored and hacktivist groups like Handala, primarily targeting the financial, ecommerce, and healthcare sectors.
The integration of AI in vulnerability research has led to a surge in false-positive bug reports, overwhelming vendors and bug bounty programs. Human oversight remains essential to validate AI findings and maintain the integrity of the CVE ecosystem.
Security researchers identified six malicious Composer packages on Packagist masquerading as OphimCMS themes. These packages contain trojanized JavaScript that executes client-side attacks, including URL exfiltration, ad injection, and redirects to gambling sites operated by the OFAC-sanctioned FUNNULL network.
The ongoing geopolitical conflict involving Iran has triggered significant cyber and influence operations, with multiple nation-state and hacktivist groups leveraging the crisis for espionage, destructive attacks, and narrative manipulation. Organizations are advised to prepare for a surge in Iranian cyber activity as domestic internet blackouts lift, alongside heightened risks of physical threats and supply chain disruptions.
A suspected China-nexus threat actor tracked as CL-STA-1087 has been conducting a persistent espionage campaign against Southeast Asian military targets since 2020. The attackers utilize custom malware, including the AppleChris and MemFun backdoors, leveraging Dead Drop Resolvers (DDR) like Pastebin and Dropbox for C2 resolution alongside advanced evasion techniques like process hollowing and DLL hijacking.
This promotional article highlights Akamai's upcoming presence at RSAC 2026, focusing on the escalating arms race between AI-driven cyber threats and enterprise security. It emphasizes that adversaries are using AI to automate API attacks and exploit cloud misconfigurations, necessitating a shift away from legacy security toward robust Zero Trust frameworks and strategic partner ecosystems.
Microsoft's March Patch Tuesday addressed 84 vulnerabilities across 15 product families, including 8 Critical and 76 Important flaws. While no zero-days were reported as actively exploited, two vulnerabilities have been publicly disclosed, and six are deemed highly likely to be exploited within 30 days. Organizations are advised to prioritize patching for critical Remote Code Execution and Elevation of Privilege vulnerabilities affecting Windows, Office, and Azure environments.
