Find what you’re looking for

Introducing Augustus: Praetorian's Open-Source LLM Vulnerability Scanner

In February 2026, Praetorian released Augustus, an open-source vulnerability scanner designed to test Large Language Models (LLMs) against a comprehensive suite of adversarial attacks. Augustus automates over 210 distinct attack vectors, including prompt injections and jailbreaks, across 28 LLM providers. This tool addresses the growing need for robust security testing as enterprises rapidly integrate generative AI into their products. By providing a portable, single-binary solution, Augustus facilitates seamless integration into continuous integration/continuous deployment (CI/CD) pipelines, enabling security teams to identify and mitigate vulnerabilities efficiently. The release of Augustus underscores the escalating threats targeting LLMs, as adversaries increasingly exploit these models for malicious purposes. The tool's comprehensive testing capabilities highlight the necessity for organizations to proactively assess and fortify their AI systems against evolving attack methodologies.

South Korea's NTS Security Lapse Results in $4.8M Crypto Theft

In February 2026, South Korea's National Tax Service (NTS) conducted raids on 124 high-value tax evaders, seizing digital assets worth approximately $5.6 million. During a press release showcasing the operation, the NTS inadvertently published images displaying a Ledger hardware wallet alongside a handwritten note containing the wallet's mnemonic recovery phrase. This exposure allowed an unauthorized individual to access and transfer 4 million Pre-Retogeum (PRTG) tokens, valued at about $4.8 million, from the confiscated wallet. The NTS has since apologized for the oversight and initiated measures to prevent similar incidents in the future. ([koreajoongangdaily.joins.com](https://koreajoongangdaily.joins.com/news/2026-03-01/national/socialAffairs/Police-probing-unauthorized-crypto-transfer-after-NTS-inadvertently-shared-wallet-recovery-phrase/2534349?utm_source=openai)) This incident underscores the critical importance of secure handling and storage of digital assets, especially by governmental agencies. As cryptocurrency adoption grows, ensuring robust security protocols and staff training is essential to prevent such costly errors and maintain public trust.

Kwamaine Jerell Ford's 2026 Phishing Scheme Targets Professional Athletes

In March 2026, Kwamaine Jerell Ford, a 34-year-old from Georgia, was indicted for orchestrating a sophisticated phishing scheme targeting professional NBA and NFL athletes. While incarcerated for a similar offense, Ford allegedly impersonated an adult film star to deceive athletes into providing their iCloud credentials and multifactor authentication codes. This access enabled him to steal sensitive personal and financial information, leading to unauthorized transactions exceeding 2,000 instances between November 2020 and September 2024. The scheme also involved coercing an OnlyFans model into recording commercial sex acts with athletes without their consent, further complicating the legal ramifications. This incident underscores the persistent threat of social engineering attacks, even from individuals previously convicted of similar crimes. It highlights the critical need for continuous vigilance, robust cybersecurity measures, and comprehensive education on recognizing and mitigating phishing attempts, especially for high-profile individuals who are frequent targets.

UK's Companies House Security Flaw Exposes Business Data - 2026

In March 2026, the UK's Companies House disclosed a significant security vulnerability in its WebFiling service, which had been present since October 2025. This flaw allowed authenticated users to access and potentially modify sensitive information of any registered company by exploiting a back-navigation loophole. The exposed data included directors' residential addresses, email addresses, and dates of birth. The agency has since rectified the issue, notified affected parties, and reported the incident to the Information Commissioner's Office (ICO) and the National Cyber Security Centre (NCSC). This incident underscores the critical importance of rigorous security testing and prompt response to vulnerabilities in public sector digital services. The exposure of personal data over an extended period raises concerns about potential misuse and the necessity for enhanced monitoring and compliance measures to protect sensitive information.

GoPIX Banking Trojan: A New Threat to Brazil's PIX Payment System

In December 2022, the GoPIX banking Trojan emerged, targeting users of Brazil's PIX instant payment system. Disguised as a WhatsApp Web installer, it spread through malicious ads, leading victims to download malware that intercepts and manipulates PIX transactions. GoPIX employs sophisticated techniques, including IP Quality Score's anti-fraud tools, to evade detection and ensure successful infections. ([usa.kaspersky.com](https://usa.kaspersky.com/about/press-releases/kaspersky-crimeware-report-reveals-new-rhysida-ransomware-lumar-stealer-and-gopix-banking-malware?utm_source=openai)) The rise of GoPIX underscores a growing trend of cybercriminals exploiting popular payment systems in Latin America. Its advanced evasion methods and focus on real-time transaction manipulation highlight the need for enhanced security measures and user awareness to combat such evolving threats. ([usa.kaspersky.com](https://usa.kaspersky.com/about/press-releases/kaspersky-crimeware-report-reveals-new-rhysida-ransomware-lumar-stealer-and-gopix-banking-malware?utm_source=openai))

ClickFix Campaigns Exploit AI Tool Installers to Deploy MacSync Infostealer

In late 2025 and early 2026, multiple ClickFix campaigns emerged, targeting macOS users with the MacSync infostealer. These campaigns utilized malicious Google Ads and AI-generated content to lure users into executing terminal commands that installed the malware. The MacSync infostealer is capable of exfiltrating credentials, browser data, and cryptocurrency wallet information. ([cybernews.com](https://cybernews.com/security/hackers-spread-mac-infostealer-using-google-ads/?utm_source=openai)) This incident underscores a growing trend of sophisticated social engineering attacks that exploit user trust in AI tools and search engine results. The increasing prevalence of such tactics highlights the need for heightened vigilance and user education to prevent similar breaches. ([techmonk.economictimes.indiatimes.com](https://techmonk.economictimes.indiatimes.com/news/security-alert/security-alert-clickfix-campaign-abuses-claude-artifacts-and-google-ads-to-drop-macos-infostealer/128334810?utm_source=openai))

PLUGGYAPE Malware: A New Cyber Threat Targeting Defense Sectors

Between October and December 2025, Ukrainian defense forces were targeted by a cyber-espionage campaign attributed to the Russian-affiliated group Void Blizzard (also known as Laundry Bear or UAC-0190). The attackers utilized messaging platforms such as Signal and WhatsApp to impersonate charitable organizations, distributing password-protected archives containing the PLUGGYAPE backdoor malware. Once executed, PLUGGYAPE enabled remote code execution, system reconnaissance, and data exfiltration, maintaining persistence through Windows Registry modifications. This campaign underscores the evolving tactics of state-sponsored actors in leveraging social engineering and trusted communication channels to infiltrate sensitive targets. The incident highlights the increasing sophistication of cyber threats facing defense sectors, emphasizing the need for enhanced vigilance and robust security measures to counteract such espionage activities.

GlassWorm Attack 2026: A Wake-Up Call for Open-Source Security

In early 2026, the GlassWorm malware campaign exploited stolen GitHub tokens to inject malicious code into numerous Python repositories. Attackers targeted projects such as Django applications, machine learning research code, Streamlit dashboards, and PyPI packages by appending obfuscated code to files like setup.py, main.py, and app.py. This code, often concealed using invisible Unicode characters, enabled the exfiltration of sensitive data, including SSH keys, cloud credentials, and cryptocurrency wallet information. The malware's command-and-control infrastructure leveraged the Solana blockchain, complicating detection and mitigation efforts. The resurgence of GlassWorm highlights the persistent vulnerabilities within software supply chains, emphasizing the need for robust security measures in open-source ecosystems. The attack underscores the importance of vigilant monitoring and the implementation of stringent access controls to prevent unauthorized code modifications and protect sensitive information.

Critical Chrome Zero-Day Vulnerability CVE-2026-2441 Patched

In February 2026, Google identified and patched a high-severity zero-day vulnerability in its Chrome browser, designated as CVE-2026-2441. This use-after-free flaw in the CSS component allowed attackers to execute arbitrary code by enticing users to visit maliciously crafted HTML pages. The vulnerability was actively exploited in the wild, prompting Google to release an emergency update to mitigate the risk. Users were urged to update their browsers immediately to versions 145.0.7632.75/76 for Windows and macOS, and 144.0.7559.75 for Linux. ([securityweek.com](https://www.securityweek.com/google-patches-first-actively-exploited-chrome-zero-day-of-2026/?utm_source=openai)) The exploitation of CVE-2026-2441 underscores the persistent threat posed by zero-day vulnerabilities and the importance of timely software updates. This incident highlights the need for organizations and individuals to maintain vigilant cybersecurity practices, including regular patching and monitoring for emerging threats.

2025 Identity Threat Landscape: The Rise of Infostealer Malware and Credential Theft

In 2025, the cybersecurity landscape witnessed an unprecedented surge in credential theft, with Recorded Future detecting 1.95 billion malware combo list credential exposures, 36 million database combo list credential exposures, 24 million database dump credential exposures, and 892 million malware log credential exposures. This escalation was primarily driven by the proliferation of infostealer malware, which harvested credentials from both personal and corporate devices, leading to significant breaches across various sectors. Notably, the Lumma Stealer emerged as the most prevalent infostealer, compromising over 394,000 Windows computers between March and May 2025. The widespread availability of stolen credentials on dark web marketplaces facilitated unauthorized access to corporate networks, resulting in data breaches, financial losses, and reputational damage for affected organizations. The current relevance of this incident is underscored by the continuous evolution of infostealer malware and the increasing sophistication of cybercriminal tactics. The commodification of credential theft has lowered the barrier to entry for attackers, enabling even less experienced individuals to execute complex attacks. This trend highlights the urgent need for organizations to adopt comprehensive identity protection strategies, including continuous monitoring, multi-factor authentication, and employee education, to mitigate the risks associated with credential-based attacks.

LiveChat Phishing Attack Exposes Sensitive User Data

In March 2026, attackers exploited the LiveChat customer support platform to impersonate reputable companies like PayPal and Amazon. They engaged victims in real-time chats, coercing them into divulging sensitive information such as account credentials, credit card details, and multifactor authentication codes. This sophisticated social engineering campaign highlights the evolving nature of phishing attacks, making them increasingly difficult to detect and prevent. The incident underscores a broader trend of cybercriminals leveraging trusted platforms to execute phishing schemes. As attackers refine their methods, organizations must enhance their security measures and user education to mitigate the risks associated with such deceptive tactics.

Protecting Cloud Infrastructure from SSRF Attacks on Proxy Servers

In March 2026, security researchers identified a series of attacks targeting misconfigured proxy servers to access cloud metadata services. Attackers exploited Server-Side Request Forgery (SSRF) vulnerabilities to send requests to internal metadata endpoints, such as 169.254.169.254, aiming to retrieve sensitive information like IAM credentials. These attacks leveraged various URL encoding techniques, including IPv4-mapped IPv6 addresses and long unsigned integer forms, to bypass security filters. The exploitation of these vulnerabilities could lead to unauthorized access to cloud resources, data exfiltration, and potential lateral movement within networks. This incident underscores the critical need for organizations to secure proxy configurations and implement robust input validation to prevent SSRF attacks. The increasing sophistication of attackers in exploiting cloud infrastructure vulnerabilities highlights the urgency for continuous monitoring and adherence to security best practices to protect sensitive data and maintain compliance with regulatory standards.