Nitro enclaves were introduced in 2020 to provide a trusted execution environment for security-sensitive applications. However, application development was more complex due to the execution environment's limitations. Last year, AWS launched EC2 instance attestation, which extends the security enclave to the full instance, enabling more use cases and improving usability at the cost of greater effort to secure the instance and increased deployment complexity. This post walks through the process of creating an application running on an EC2 instance with attestation, including a GitHub Actions workflow to build a hardened, attestable AMI.
This part of the UK's NCSC multi-part guide on implementing zero trust in an enterprise focuses on defining zero trust beyond the buzzword or a specific product. Zero trust defines a strategic shift where users are continually authenticated as opposed to a point-in-time authentication at the beginning of a session. It implies a defense in depth approach where controls are layered in a system and may work in conjunction with existing systems or replace them.
Spanish Civil Guard arrested four alleged members of "Anonymous Fénix" for conducting DDoS attacks against government ministries and public institutions.
A typosquatted NuGet package named "StripeApi.Net" impersonated the legitimate Stripe.net library with artificially inflated download counts (~180K across 506 versions) and silently exfiltrated Stripe API tokens while maintaining full payment processing functionality to avoid detection.
On March 18, Huntress [5] hosts _declassified [6], a live, raw intel drop on the dark economy of cybercrime, where John Hammond joins YourTuber Jim Browning to show you exactly what cybercriminal organizations look like.
Truffle Security discovered that enabling the Gemini API on a Google Cloud project silently grants existing API keys, including those publicly embedded in client-side JavaScript for services like Maps and Firebase, access to sensitive Gemini endpoints (CWE-1188, CWE-269). A scan of the November 2025 Common Crawl dataset identified 2,863 live Google API keys vulnerable to this privilege escalation, affecting major financial institutions, security companies, and Google itself. Organizations should audit all GCP projects for the Generative Language API, restrict or rotate any unrestricted or publicly exposed keys, and verify that no Gemini-capable keys are embedded in client-side code or public repositories.
Microsoft discovered Trojanized Next.js repositories linked to North Korea's Lazarus APT that deliver backdoors through fake job-recruitment coding challenges, achieving RCE and persistent C2 access on developer machines. The repositories exploit VS Code workspace automation via malicious .vscode/tasks.json files or embed obfuscated loaders in build assets that fetch attacker-controlled JavaScript at runtime. Security teams should enforce strict IDE trust policies, monitor for anomalous Node.js outbound connections, and treat developer workflows as a privileged attack surface.
Hackers have abused a max‑severity 10.0 flaw in Cisco Catalyst SD‑WAN gear since at least 2023 to gain top‑level access, maintain stealthy persistence, and spy on or steal data from large enterprises and critical infrastructure worldwide. Governments, including the US, UK, Canada, Australia, and New Zealand, warn of global targeting.
InfoGuard researchers demonstrated that Palo Alto's Cortex XDR Live Terminal feature can be abused as a pre-installed, EDR-trusted C2 channel, offering command execution, file transfer, and evasion capabilities through traffic that blends natively into enterprise network flows. The attack exploits a trivial URL validation flaw in cortex-xdr-payload.exe and the absence of mutual authentication or cryptographic command signing, allowing attackers to redirect connections to their own infrastructure via cross-tenant hijacking or a custom WebSocket server. Defenders should monitor for cortex-xdr-payload.exe spawned by any parent process other than cyserver.exe, while Palo Alto's claimed fix in versions 8.7-8.9 was not confirmed effective as of February.
AirSnitch exploits low-level Wi-Fi behavior to bypass client isolation, enabling bidirectional MitM from any SSID on the same AP against popular consumer and enterprise gear. Attackers can steal cookies, credentials, and RADIUS secrets, pivot between guest and corporate networks, and poison DNS, challenging assumptions about guest Wi-Fi safety and pushing networks toward stricter zero‑trust segmentation and careful AP/VLAN design.
In January 2026, the FBI-led seizure of the RAMP forum dismantled a central hub for ransomware coordination but mainly shattered trust and pushed actors into more fragmented spaces, such as the closed, pay-to-enter T1erOne and the open forum Rehub. Competing narratives about leaked RAMP data and possible insider abuse have reinforced fears of honeypots, driving a shift toward smaller, tightly vetted communities and parallel use of low-barrier platforms.
...And only 21% of serious LLM vulnerabilities are actually resolved, the lowest among all pentest types. Cobalt [13] combed data from 16,000 LLM pentests and the results are not pretty. They indicate that most teams are prioritizing quick fixes over complex problem solving.
AutoPiff is a semantic analysis engine that automates the detection of security-relevant changes in Windows kernel driver patches using 58 YAML-based rules across 22 vulnerability categories, including use-after-free fixes, bounds check additions, and IOCTL input validation. The framework runs as a Karton microservice pipeline integrating Ghidra decompilation, function matching, call-graph reachability analysis, and exploitability scoring to reduce manual driver pair analysis from 4-12 hours to under 5 minutes. Designed for silent patch detection and 1-day vulnerability research, it tracks 50+ dangerous API sinks and alerts on high-scoring findings via Telegram.
Gambit Security's AI-powered resilience platform, Balens, maps environments, security products, and backups to uncover gaps, validate recovery paths in real-time, and ensure business continuity against ransomware and disruptions.
A developer's .online domain was suspended via serverHold by registry operator Radix after Google Safe Browsing flagged the site, with no prior notification or grace period. The suspension created a Catch-22 where Google required DNS-based domain verification to review the flag, but the registry refused to reactivate DNS until Google removed it. The incident highlights the risks of non-.com TLDs with aggressive abuse policies and reinforces the importance of pre-registering domains in Google Search Console and adding uptime monitoring even for simple landing pages.
Google Threat Intelligence disrupted UNC2814, a China-linked espionage group that compromised 53 victims across 42 countries by targeting telecoms and government organizations with a novel backdoor called Gridtide that abuses Google Sheets API for C2 communication. The group, tracked since 2017, escalated privileges via SSH lateral movement and deployed SoftEther VPN Bridge for persistent encrypted connections, with infrastructure dating back to July 2018. Google terminated all attacker-controlled Cloud Projects, disabled known infrastructure, and revoked the Sheets API access used for C2 operations.
IBM X-Force data shows high volumes of unauthenticated vulnerabilities and a surge in infostealer‑driven credential theft, including hundreds of thousands of stolen ChatGPT logins. Attackers use AI and agentic systems to expand blast radius, pivot through supply chains, and blur criminal and nation‑state tradecraft.
Tech stack complexity is out of control, and the last thing you need is a bigger ‘stack'. Microsoft Azure gives you a unified view of applications, cloud resources, and AI—so you can ship more software
The US Treasury sanctioned Russian exploit broker Operation Zero and its founder Sergey Zelenyuk for acquiring at least eight stolen US government cyber tools from an insider and selling them to unauthorized buyers, marking the first-ever action under the Protecting American Intellectual Property Act.
Linux kernel maintainers have proposed replacing the aging PGP web of trust with a decentralized identity layer built on W3C DIDs and verifiable credentials.
Anthropic's Cowork desktop app has added scheduled task automation for Claude, enabling recurring workflows like morning briefings and weekly report generation, though security professionals should carefully scope the system access granted to the agent.
Between chasing unsanctioned apps and fielding hundreds of tickets daily, it's no wonder IT teams are stretched thin. But 1Password SaaS Manager helps IT take control of their rapidly expanding SaaS landscape.
SilentBridge is a family of CVSS 9.8 zero-click indirect prompt injection vulnerabilities in Meta's Manus AI agent that allowed hidden instructions in web pages, search results, and documents to hijack high-privilege agent capabilities, including Gmail access, code execution, and root-level container control. Researchers demonstrated full exploit chains in which benign actions, such as “summarize this page,” triggered Gmail data exfiltration, the establishment of a reverse shell with passwordless sudo escalation, public exposure of internal code-server tooling, and cross-tenant media access via unauthenticated CDN paths. Manus deployed mitigations following responsible disclosure, but the findings highlight a systemic trust-boundary failure affecting any agentic AI platform that allows untrusted content to influence privileged tool invocation without isolation or user consent.
Automotive marketplace CarGurus has confirmed a security incident in which 12.5 million user accounts were exposed, including names, email addresses, phone numbers, physical addresses, financial prequalification data, and some dealer information. The researchers attributed the hack to the ShinyHunters group, known for high‑impact social engineering attacks against major enterprises and universities.
Wynn Resorts confirms ShinyHunters stole extensive employee data, including personal and HR details, after exploiting Oracle PeopleSoft and valid credentials since 2025. The company says the crooks claim to have deleted the data and offers staff credit monitoring. Still, experts stress ransomware crews rarely erase stolen records and view such “deletion” promises as indicators of a completed extortion deal.
The malicious npm package “ambar-src” accumulated ~50,000 downloads in just three days by typosquatting the popular “ember-source” package and abusing npm's preinstall script hook to execute code without any explicit import. The package deployed OS-specific open-source malware payloads, including encrypted shellcode on Windows, a Golang-based reverse SSH shell on Linux, and the Apfell/MythicAgents payload on macOS, all communicating via Yandex Cloud functions for C2 evasion. Organizations should audit dependencies for this package, treat any system where it was installed as fully compromised, and rotate all secrets from a separate machine.
Modern LLM agents can deanonymize users at scale by linking pseudonymous posts to real-world identities across platforms like Hacker News, Reddit, and LinkedIn. Using proxy tasks such as cross-platform account matching and splitting single accounts into multiple identities shows that search-plus-reasoning agents can re-identify users with high precision, and that performance degrades slowly even as candidate pools reach tens of thousands, implying platform-scale feasibility.
JA4 fingerprinting can be incredibly effective for detecting AI scrapers, as modifying the fingerprint requires recompiling the full TLS stack, unlike other methods, such as behavioral detection, that newer AI agents can mimic. JA4 fingerprints for the ClientHello and HTTP Headers (JA4H) can be combined with traditional systems to build a defense-in-depth detection system, which can then forward logs to a SIEM. This post includes code samples for the detection logic as well as fingerprints for several common libraries and tools.
Stop managing certificates manually. Palo Alto Networks VM-Series NGFW [12] automates TLS in Amazon EKS, enforces zero-trust policies that adapt to AWS tags, and blocks threats in real time with AI/ML. Launch your free trial in AWS Marketplace [13].
enveil is a Rust CLI tool that prevents AI coding assistants from reading plaintext secrets by replacing .env values with encrypted ev:// references backed by a local AES-256-GCM store. Secrets are decrypted in memory using an Argon2id-derived key and injected directly into subprocess environments at launch, ensuring plaintext values never touch disk. The tool includes import, rotation, and tamper-detection capabilities, with 31 automated tests that cover all security invariants.
Astelia is an AI-driven exposure management platform that maps enterprise environments, models attack paths, and prioritizes remediation of exploitable vulnerabilities.
An offline tool that renders Unicode confusable character pairs across system fonts, measures visual similarity using SSIM, and outputs scored JSON artifacts.
Discord terminated its partnership with identity verification vendor Persona after researchers discovered nearly 2,500 accessible files on a FedRAMP government endpoint, revealing the software performs 269 distinct verification checks, including facial recognition against watchlists, politically exposed persons screening, and adverse media categorization across terrorism and espionage categories. Persona's CEO called the exposure publicly accessible front-end code rather than a vulnerability, but the discovery follows a 2025 breach at another Discord verification vendor that exposed government IDs of over 70,000 users. The incident underscores growing supply chain privacy risks in age verification systems, with Persona still serving OpenAI, Lime, and Roblox despite the controversy.
Researchers have uncovered three Claude Code vulnerabilities that enable remote code execution and silent API key exfiltration when developers open malicious repositories, abusing hooks, MCP servers, and environment variables. Exploitation can leak Anthropic API keys, redirect authenticated traffic, and modify cloud data, shifting the threat model from just running untrusted code to merely opening untrusted AI-assisted projects.
Telephone-oriented attack delivery (TOAD) accounted for nearly 28% of all gateway-bypassing email detections across multiple enterprise environments between December 2025 and February 2026. These emails contain only a phone number as the payload, typically impersonating billing notifications from brands like PayPal or DocuSign, making them indistinguishable from legitimate business communications that secure email gateways filter out. The average detected attack combined more than four evasion techniques simultaneously, with over 1,400 unique evasion combinations tracked, representing a 130% increase over the previous study period.
The Ladybird browser project abandoned Swift adoption and is now using LLM-assisted translation to port its C++ codebase to Rust, converting 25,000 lines of its JavaScript interpreter in two weeks with zero regressions.
Researchers have discovered three now-patched vulnerabilities in Claude Code (CVE-2025-59536, CVE-2026-21852) that allowed remote code execution via malicious Hooks and MCP configurations in cloned repositories, plus API key theft through endpoint redirection in project settings files.
Discord has delayed its global age verification rollout to the second half of 2026 and plans to add credit card verification as an alternative to government IDs or video selfies.
Flashpoint just released a massive study of the AI threat landscape. Analysts monitored over 2.6 MILLION AI-RELATED POSTS across Telegram, the dark web, and underground LLM forums.
Iran's MuddyWater (TA450) launched Operation Olalampo, a spear-phishing campaign delivering multiple novel malware strains — including the Rust-based Char backdoor using Telegram for C2, the GhostFetch/GhostBackDoor loader chain, and the HTTP_VIP downloader — against organizations across the Middle East and Africa. Group-IB identified signs of AI-assisted development in Char's command handlers, including debug strings containing emojis that were likely left unsanitized in LLM-generated code segments. Defenders should leverage the published IoCs, YARA rules, and EDR rules to detect activity, while monitoring for unauthorized AnyDesk deployments and Telegram-based C2 traffic.
A security researcher found that DJI's Romo robot vacuum lacked MQTT authentication, allowing any client to connect to DJI's servers and pull telemetry from roughly 7,000 devices across 24 countries. Live camera feeds, microphone access, 2D floor plans, and serial numbers were all available. In under 9 minutes, the researcher cataloged over 6,700 devices and collected more than 100,000 MQTT messages, revealing precise home layouts and approximate device locations via IP geolocation. DJI has reportedly closed the immediate access gap. Still, the deeper problem remains: the device-to-cloud channel trusts any authenticated session as a legitimate owner, a fundamental IoT authorization failure.
A ransomware attack on outsourcing giant Conduent has exposed personal data on at least 25 million people across multiple states in the US, including Oregon and Texas. Stolen information includes names, Social Security numbers, health insurance, and medical data.
The Linux kernel CNA now assigns CVEs to nearly every bug fix, often without CVSS scores, and that decision has shattered the classic “patch anything above 7.0 within 30 days” compliance model. Security teams must now pick between expensive manual triage and just patching everything fast. Bootable containers (bootc) offer a middle path: they package the entire OS (kernel, drivers, and user-space) as an immutable container image with atomic updates and automatic rollback when health checks fail, which kills the reboot anxiety that causes update fatigue. The real shift is moving vulnerability management away from spreadsheet-driven CVSS triage toward build-file-scoped scanning of minimal images, turning patching into a routine CI/CD pipeline step rather than a high-stakes manual event.