This is the first post in a series using BishopFox's BadPods, which features a variety of pod manifests with varying levels of misconfiguration, to investigate the protections included in different cloud providers' managed Kubernetes offerings. This series includes the most permissive pod manifest, which the author uses to trivially escape the pod via a mounted host path and then further escalate privileges within the account by stealing IAM credentials from IMDS.
BinaryAudit is an open-source benchmark from Quesma for evaluating AI agents' ability to detect backdoors in stripped binary executables using reverse engineering tools like Ghidra and Radare2. The benchmark tests models against backdoored versions of lighttpd, dnsmasq, Dropbear, and Sozu, with Claude Opus 4.6 leading at 49% detection but hampered by a 22% false positive rate across clean binaries. Results highlight that while AI-assisted binary analysis is becoming accessible to non-specialists, high false positive rates and inability to trace data flow through complex binaries make it unsuitable for production malware detection today.
Nucleus provides an exposure management orchestration platform that ingests data from 200+ security and asset tools, normalizes and correlates vulnerabilities with threat intel and business context to prioritize and remediate risk across environments.
Dotenv is a zero-dependency module that loads environment variables from a .env file into process.env, storing configuration in the environment and separating from code while following the Twelve-Factor App methodology.
The Kimwolf IoT botnet accidentally disrupted the I2P anonymity network after attempting to register 700,000 infected devices as nodes — a Sybil attack that overwhelmed the roughly 15,000-20,000 node network — while seeking takedown-resistant C2 infrastructure. The botmasters, who admitted to the disruption on Discord, have been experimenting with both I2P and Tor as fallback command-and-control channels after security companies and network operators intensified efforts to combat the botnet. I2P is currently operating at roughly half capacity as stability fixes roll out, while Kimwolf's overall numbers have reportedly dropped by more than 600,000 infected systems following internal operator disputes and operational errors.
Microsoft has shipped emergency fixes for several actively exploited Windows and Office zero-days that enable one-click remote code execution and SmartScreen bypasses across all supported Windows versions. Attackers can silently run high-privilege malware, enabling ransomware and espionage campaigns. Admins and end users are urged to patch immediately and treat malicious links and Office files as high risk.
Google's threat team details how North Korean group UNC2970 and others weaponize Gemini for OSINT-driven target profiling, tailored phishing, vulnerability analysis, and exploit debugging, blurring the line between research and reconnaissance. Attackers also use Gemini's API in HONESTCUE malware, AI-built COINBAIT phishing kits, and large-scale model extraction attempts that clone Gemini's behavior via high-volume prompt queries.
Law enforcement crippled Lumma Stealer's infrastructure in 2025, but the infostealer has rapidly rebuilt and is spreading again using persuasive “ClickFix” fake CAPTCHA prompts that trick users into running commands in Windows Terminal, loading CastleLoader in memory, then Lumma.
Developer tool sprawl is real. You've got one tab for logs, another for metrics, half a dozen for cloud services, and somewhere in there is the code you're actually trying to ship.
A new commercial mobile spyware toolkit, ZeroDayRAT, sold via Telegram and analyzed by iVerify, offers nation-state-level capabilities, including live camera and microphone feeds, keylogging, GPS tracking, bank credential theft, and crypto clipboard hijacking, for both Android and iOS devices. The self-hosted kit uses a builder model in which operators generate payloads for their own infrastructure, distributing them via phishing, smishing, or trojanized apps. The creators deliberately obscure attribution by advertising in five languages and mixing Chinese, Russian, and Indian targeting indicators. Takedown efforts have been complicated by the decentralized operator model and Telegram's slow enforcement, making this a persistent threat with few available IoCs beyond unusual battery drain and unexplained financial transactions.
Unknown threat actors exploited vulnerable SolarWinds Web Help Desk instances in December 2025 to gain initial access, with Microsoft unable to confirm whether the attacks leveraged CVE-2025-40551 (9.8 CVSS deserialization RCE), CVE-2025-40536 (8.1 CVSS auth bypass), or CVE-2025-26399 (9.8 CVSS command execution). Post-exploitation activity included BITS abuse for payload delivery, installation of Zoho ManageEngine RMM for persistence, LSASS credential dumping via DLL sideloading, and DCSync attacks to extract domain controller password data. Organizations should immediately patch WHD, remove public access to admin paths, scan for unauthorized RMM tools such as ToolsIQ.exe, and rotate service and admin account credentials accessible from WHD.
Koi Security's threat hunting agent analyzed skills being offered on OpenClaw's ClawHub marketplace and discovered that 341/2857 of them were malicious. Most malicious extensions follow a similar pattern: the prerequisites instruct the agent to download and execute a file containing a malicious payload that installs the AMOS stealer. The post also highlights a few unique skills that download other payloads or hide their stealers and backdoors within the code itself.
The collapses of Black Basta and LockBit have triggered a violent restructuring of the RaaS ecosystem, with new entrants like DragonForce operating as vertically integrated "cartels" that embed access broker marketplaces directly into affiliate panels, compressing the window between initial access and ransomware deployment to hours. A parallel commoditization of EDR evasion tools, including BYOVD-based killers exploiting a still-valid 2006 Microsoft-signed driver, has collapsed the barrier to entry so that even low-budget operators can defeat enterprise security stacks. Defenders should prioritize enabling Microsoft's opt-in Vulnerable Driver Blocklist, monitor access broker marketplaces for organizational exposure, and shift threat intelligence from tracking ransomware brands to tracking persistent infrastructure indicators like shared SSH fingerprints and certificate reuse.
VoidLink is a Zig-based Linux C2 implant that bears strong indicators of LLM-generated code — including structured "Phase X:" labels, verbose debug logging, and documentation patterns left in the production binary — demonstrating how AI coding agents are lowering the barrier to producing functional, multi-cloud malware. The implant fingerprints AWS, GCP, Azure, Alibaba Cloud, and Tencent Cloud environments, harvests credentials from environment variables and metadata APIs, performs container escape via Docker/Kubernetes plugins, and deploys an adaptive kernel-level rootkit that selects eBPF, LKM, or LD_PRELOAD stealth based on host kernel version. Defenders should monitor for unauthorized cloud metadata API queries from workloads, audit Kubernetes service account token access, and treat verbose or structured debug artifacts in captured malware as potential indicators of AI-assisted development requiring adjusted threat modeling.
Attackers have shifted from credential theft to stealing authentication tokens for initial access. In one incident that the author observed, an attacker successfully stole an authentication token via an attacker-in-the-middle (AitM) Microsoft 365 proxy, but couldn't achieve their goals due to rapid detection and response. In a second incident, the attacker was unable to steal a token because the organization utilized FIDO2 security keys, so the authentication attempt through the AitM proxy failed.
Every new monitoring tool, service, or "quick integration" adds another dashboard to the pile—and before you know it, you're spending more time switching between tools than actually building. With Microsoft Azure [12], your entire stack lives in one workspace: infrastructure, data, AI, and more. See how Azure brings it all
Skill Scanner is a security scanner for AI agent skills that detects prompt injection, data exfiltration, and malicious code patterns via a combination of pattern-based detections, LLM-as-a-judge, and behavioral data analysis.
Mandiant attributed an intrusion against a FinTech entity to UNC1069, a DPRK-nexus financially motivated threat actor that leveraged a compromised Telegram account, a spoofed Zoom meeting with a reported deepfake video, and a ClickFix attack to deploy seven malware families on a macOS device — including new tools SILENCELIFT, DEEPBREATH, and CHROMEPUSH. The attack chain progressed from the WAVESHAPER backdoor through the HYPERCALL downloader to deploy HIDDENCALL for hands-on keyboard access, while DEEPBREATH bypassed macOS TCC protections to steal Keychain credentials, browser data, and Telegram sessions, and CHROMEPUSH installed a malicious Chrome extension for keylogging and cookie theft. The investigation highlights UNC1069's expanding use of GenAI tools like Gemini and GPT-4o for social engineering, alongside a significant tooling refresh targeting cryptocurrency startups, software developers, and venture capital firms.
The certificates that Microsoft uses for Secure Boot are set to expire in June 2026 after 15 years of use. Users who have automatic updates enabled will receive the new updates automatically. Admins can deploy the new certificates via registry keys, Group Policy, or Windows Configuration System (WinCS). Windows 10 users who are not enrolled in extended security updates will not receive the new certificates.
Reynolds ransomware bundles a vulnerable NsecSoft NSecKrnl driver (CVE-2025-68947, CVSS 5.7) directly in its payload to terminate EDR processes from CrowdStrike, Sophos, Symantec, Palo Alto Cortex XDR, and Avast before encrypting files, eliminating the need for a separate BYOVD deployment step. Symantec and Carbon Black researchers noted a suspicious side-loaded loader on the target network weeks before ransomware execution, followed by deployment of the GotoHTTP remote access tool for persistent access. The article also covers broader ransomware developments, including LockBit 5.0's shift to ChaCha20 encryption with wiper capabilities, Interlock's use of a gaming driver zero-day (CVE-2025-61155) for BYOVD attacks, and ransomware actors increasingly targeting misconfigured AWS S3 buckets.
"Barriers to Secure OT Communications: Why Johnny Can't Authenticate" is a guide that addresses the cost, complexity, latency, and interoperability barriers preventing critical infrastructure operators from adopting secure industrial protocols.
Engineering and security teams often rely on tools that reveal only a narrow slice of what is happening inside their software. The deeper risks in modern code such as logic failures, unexpected execution paths, and behavioral deviations from intent rarely appear without more advanced analysis.
Fortinet has patched CVE-2026-21643, a critical SQL injection vulnerability in the FortiClientEMS GUI that allows unauthenticated remote attackers to execute arbitrary code or commands by improperly neutralizing special characters in SQL queries. The flaw affects FortiClientEMS version 7.4.4. Version 7.4.5 was released as a fix. Versions 7.2 and 8.0 are unaffected. While no exploitation in the wild or public proof-of-concept has been observed yet, organizations should patch urgently, given Fortinet's history as a heavily targeted vendor per CISA's Known Exploited Vulnerabilities Catalog.
Forcepoint X-Labs discovered the Phorpiex botnet delivering Global Group ransomware, a Mamona successor, via phishing emails containing .lnk files disguised as documents. It leverages living-off-the-land techniques using PowerShell and Command Prompt for payload delivery. The ransomware operates in a mute mode that generates encryption keys locally with ChaCha20-Poly1305 instead of contacting a C2 server, allowing it to encrypt files on offline machines and evade network-based detection. After encryption, the malware self-deletes, destroys Volume Shadow Copies, and appends a .reco extension to locked files, leaving minimal forensic artifacts.
Researchers at the security firm LayerX uncovered a critical (CVSS 10/10) vulnerability in Claude Desktop Extensions (DXT) that could enable remote code execution (RCE) with no user interaction. In the proof-of-concept disclosed by LayerX, an attacker would send the victim a Google Calendar invite containing malicious instructions in the event description. The victim would then, at some point, ask Claude to manage their calendar or check events, and the malicious instructions would be executed. This vulnerability is enabled by the fact that Claude DXT intentionally runs without any isolation.
A security researcher discovered that the DoJ's latest Epstein document dump inadvertently included unredacted email attachments as raw base64-encoded content — left uncensored because the assigned intern didn't recognize the significance of pages of encoded data embedded in printed-and-scanned emails. Reconstructing the original PDFs proved extremely difficult due to poor OCR quality on Courier New font rendering (where "1" and "l" are nearly indistinguishable), with multiple commercial and open-source OCR tools failing until a community-developed pixel-level template-matching approach achieved 99.96% line accuracy and successfully decompressed 39 of 40 FlateDecode streams. The incident highlights how encoding artifacts and document processing pipelines can inadvertently preserve sensitive content that redaction workflows miss, and serves as a cautionary tale for forensic document handling and redaction verification processes.
In November 2025, the Sysdig Threat Research Team (TRT) detected an account compromise that was notable for the attacker's 8-minute progression from initial access to administrative privileges, using LLM assistance. The attacker gained initial access through credentials discovered in a public S3 bucket. They then used these credentials to replace the Lambda function code to escalate privileges, move laterally across 19 different principals, and finally harvest sensitive data. They also used Bedrock to access cloud-hosted LLMs and launch GPU instances for model training. The article includes mitigation recommendations for each step in the attacker's exploit chain.
AI lets developers ship in seconds, and now security can move just as fast. depthfirst [11] understands your code, business logic, and infrastructure to trace full attack paths, validate what's actually exploitable, and tell your devs exactly how to fix it. Trusted by Lovable, Supabase, Moveworks, and Persona. Book a demo and try it for
Augustus is an open-source Go-based LLM vulnerability scanner that tests models against 210+ adversarial attacks across 47 categories, including jailbreaks, prompt injection, encoding exploits, data extraction, and agent manipulation, shipping as a single binary with support for 28 LLM providers. Inspired by NVIDIA's garak but reimplemented in Go for faster execution and zero runtime dependencies, Augustus features a pipeline architecture with composable probes, buff transformations (encoding, paraphrase, poetry, and low-resource language translation), and 90+ detectors, including LLM-as-a-judge and HarmJudge. The tool is part of Praetorian's "The 12 Caesars" campaign.
Klint is a single-binary, zero-dependency incident response scanner written in C++23 that detects hidden kernel modules, rootkits, compromised syscall tables, and eBPF-based persistence mechanisms across 8 self-registering scanners. The tool cross-references multiple kernel data sources — including /proc/kcore, /proc/kallsyms, MSRs, and cgroup task files — to identify inconsistencies indicative of tampering, with structured JSON output and meaningful exit codes designed for CI pipeline integration. Builds as a fully static binary for rapid deployment via SCP into compromised environments during live incident response.
Germany's BSI and BfV warned that state-backed hackers are targeting military leaders, diplomats, and journalists across Europe by abusing Signal's legitimate features — impersonating Signal Security Support to steal PINs and hijack accounts, or tricking targets into scanning QR codes that silently link attacker devices with access to 45 days of chat history. The campaigns rely entirely on social engineering rather than exploits, making them difficult to detect through traditional security tooling. Users should enable Registration Lock, audit linked devices regularly, and treat any unsolicited messages requesting verification codes as malicious.
A hacktivist going by the name wikkid has scraped and released over 500K customer records from stalkerware services provided by a Ukrainian company called Struktura. The scraped data includes customer email addresses, which app or brand the customer paid for, how much they paid, the payment type, and the last four digits of credit card numbers. The hacktivist stated that they exploited a “trivial bug” in the service's website.
Discord has announced that beginning in early March, users will need to verify their age via a selfie or government ID. Users' accounts will be set to teen accounts by default until they perform the verification. Discord states that the selfie processing will happen locally on users' devices, that government IDs will be deleted by their third-party processors “quickly,” and that age verification status will not be visible on users' profiles.
Two Connecticut men were indicted on 45 federal counts for allegedly using ~3,000 stolen identities purchased from darknet markets to create fraudulent accounts on FanDuel, DraftKings, and BetMGM, exploiting new-user promotional bonuses to steal $3 million over five years.
Google Research used Gemini's iterative self-correction prompting to discover a bug in a cryptography paper on SNARGs for NP from LWE that human reviewers had missed, highlighting LLMs' growing potential as adversarial reviewers in academic research and security audits.