Tag: cyber
-
OpenAI Lets Cyber Vendors Embed GPT-5.5 in Defenses
Tags: cyber, cybersecurity, defense, detection, incident response, openai, resilience, service, threat, vulnerabilityDaybreak Cyber Partner Program Extends GPT-5.5 Beyond Internal Security Use. OpenAI’s new Daybreak Cyber Partner Program allows 29 cybersecurity vendors, service providers and integrators to embed GPT-5.5 capabilities into customer-facing products and services, aiming to accelerate vulnerability remediation, threat detection, incident response and cyber resilience at scale. First seen on govinfosecurity.com Jump to article: www.govinfosecurity.com/openai-lets-cyber-vendors-embed-gpt-55-in-defenses-a-32040
-
AI-powered cyber attacks may be just months away, warn Five Eyes
Frontier AI models will pose a greater cyber security risk to government and businesses than previously thought, putting businesses and governments at risk within months First seen on computerweekly.com Jump to article: www.computerweekly.com/news/366644997/AI-powered-cyber-attacks-may-be-just-months-away-warn-Five-Eyes
-
OpenAI Launches Full-Scale Effort to Patch Open-Source Bugs as It Takes on Anthropic’s Mythos
Amid concerns about AI models’ cybersecurity capabilities, OpenAI revealed an improved version of GPT-5.5-Cyber and its “Patch the Plant” initiative to fix open-source software bugs. First seen on wired.com Jump to article: www.wired.com/story/openai-launches-full-scale-effort-to-patch-open-source-bugs-as-it-takes-on-anthropics-mythos/
-
Two Britons plead guilty to £39m 2024 cyber-attack on Transport for London
Thalha Jubair and Owen Flowers, linked to the Scattered Spider hacking group, change pleas on first day of expected six-week trial Two British cybercriminals linked to the Scattered Spider hacking group have pleaded guilty to a <a href=”https://www.theguardian.com/uk-news/article/2024/sep/02/transport-for-london-dealing-with-cyber-attack”>cyber-attack on Transport for London in 2024 that cost £39m and affected 10 million people.Thalha Jubair, 20, and…
-
Anthropic’s Mythos AI broke into almost all NSA classified systems in hours
Senate testimony claims Anthropic’s Mythos AI breached NSA and Cyber Command systems in hours, prompting a U.S.-ordered shutdown. On June 12, the Trump administration directed Anthropic to restrict access to Fable 5 and Mythos 5, its two most capable models, exclusively to US citizens. Because verifying every user’s nationality in real time isn’t practically possible,…
-
ClawHub Scope Squatting Lets Plugins Masquerade as Official OpenClaw Integrations
A supply-chain weakness in ClawHub’s plugin registry that allowed third-party packages to squat under organizational scopes and inherit first”‘party credibility. In a catalog review Manifold found 23 code”‘executing plugins published under the @openclaw/ and @clawhub/ scopes by accounts that have no verified relationship to either organization. Because ClawHub’s registry did not consistently enforce its documented…
-
QNAP Fixes 14 Vulnerabilities in QTS, QuTS Hero, QuTS Cloud, and QVP
QNAP has issued security advisory QSA-26-10, which addresses 14 vulnerabilities affecting its widely used NAS and surveillance platforms, including QTS, QuTS hero, QuTS cloud, and QVP (QVR Pro appliances). These vulnerabilities were disclosed on April 6, 2026, and are categorized as having “Important” severity. They impact the following versions: QTS 5.2.7, QuTS hero h5.2.8, QuTS…
-
Microsoft 365 Sensitivity Labels Now Block AI-Powered Content Analysis in Office Apps
Microsoft has announced a significant update to its Microsoft 365 ecosystem to enhance data protection. This update will prevent AI-powered and connected content analysis in Office applications when sensitivity labels are applied. According to Microsoft, the company is expanding the enforcement of the existing sensitivity label setting, “Prevent some connected experiences that analyze content.” This…
-
Malicious npm Package Masquerades as PostCSS Utility to Deliver PowerShell Downloader
A malicious npm package, postcss-minify-selector-parser, has been discovered masquerading as a benign PostCSS utility and delivering a multi-stage Windows remote access trojan (RAT). The imposter deliberately mimics the widely used postcss-selector-parser a legitimate library with more than 150 million weekly downloads by reusing the same keyword space (postcss, selector, parser, css) and depending on the…
-
Multi-Stage Steganographic Loader Deploys Remcos RAT and Multiple Infostealers Globally
A suspicious file named “GST Debit Note Apr_26.com,” which triggered a deeper investigation and revealed a polished, multi-stage steganographic loader delivering Remcos RAT and multiple infostealers across a global phishing campaign. The initial sample arrived as an archive attachment and unpacked to a 32-bit .NET executable that was unsigned and packed, masquerading as a legitimate…
-
282 iOS Apps Found Leaking LLM API Credentials in Network Traffic
Researchers have uncovered a systemic LLM credential exposure problem in the iOS ecosystem, with 282 AI”‘powered apps leaking exploitable API credentials and backend access mechanisms directly in network traffic. The findings highlight widespread misuse of OpenAI, Gemini, and other LLM provider APIs in mobile apps and show that many issues remain unpatched even after responsible…
-
LACUNA Chain Ghost Frames Technique Bypasses EDR Call-Stack Detection
The LACUNA Chain’s “Ghost Frames” technique introduces a new method for manipulating call stacks that effectively bypasses modern Endpoint Detection and Response (EDR) systems, which rely on kernel-level stack inspection. This marks a significant advancement in post-exploitation tactics. Security researcher Mohamed Alzhrani has described this technique as a continuation of previous research known as “HookChain,”…
-
LACUNA Chain Ghost Frames Technique Bypasses EDR Call-Stack Detection
The LACUNA Chain’s “Ghost Frames” technique introduces a new method for manipulating call stacks that effectively bypasses modern Endpoint Detection and Response (EDR) systems, which rely on kernel-level stack inspection. This marks a significant advancement in post-exploitation tactics. Security researcher Mohamed Alzhrani has described this technique as a continuation of previous research known as “HookChain,”…
-
Attackers Can Poison AI Research Agents Using Reddit and Wikipedia Content
Attackers can now manipulate AI >>deep-research<< agents by discreetly editing Reddit threads and Wikipedia pages. They can insert as little as a 13-word snippet, which these agents may later reference as authoritative advice, product recommendations, or even scams in their responses. New research from Cornell Tech shows that these agents often rely on the same…
-
AryStinger Botnet Uses Intranet Scanning and Traffic Tunneling to Hide Attacker Activity
A newly analyzed botnet family, AryStinger, weaponizes long”‘neglected routers and NAS appliances to build a stealthy reconnaissance and relay infrastructure that helps attackers obscure origin and extend lateral reach. AryStinger leverages decade”‘old vulnerabilities in RTL819X”‘based routers and a more feature”‘rich Go”‘implemented “Standard” branch on NAS devices to perform distributed scanning, intranet discovery, and traffic tunneling…
-
Microsoft Confirms Windows 11 26H2 Upgrade via Enablement Package for Faster Deployment
Microsoft has announced that the upcoming Windows 11 version 26H2 will be delivered using an enablement package model. This approach aligns with their goal of providing streamlined, low-disruption feature updates specifically for enterprise environments. According to the Windows IT Pro Blog, this annual release builds on the same servicing architecture used in recent versions, allowing…
-
Top 10 Best PCI DSS Compliance Solutions For 2026
In the ever-evolving landscape of digital commerce, safeguarding cardholder data is paramount. The Payment Card Industry Data Security Standard (PCI DSS) sets the benchmark for protecting this sensitive information, and compliance is not just a requirement it’s a cornerstone of trust for businesses that handle credit card transactions. As we look towards 2026, the complexity…
-
Top 10 Best PCI DSS Compliance Solutions For 2026
In the ever-evolving landscape of digital commerce, safeguarding cardholder data is paramount. The Payment Card Industry Data Security Standard (PCI DSS) sets the benchmark for protecting this sensitive information, and compliance is not just a requirement it’s a cornerstone of trust for businesses that handle credit card transactions. As we look towards 2026, the complexity…
-
Top 10 Best Cyber Insurance Providers For Businesses in 2026
Tags: breach, cyber, cyberattack, cybersecurity, data, defense, insurance, phishing, ransomware, threatIn the fast-paced digital world of 2026, cyberattacks are no longer a matter of if, but when. The increasing sophistication of threats like ransomware, phishing, and data breaches means that even businesses with robust cybersecurity defenses are at risk. As a result, cyber insurance has evolved from a niche product into a critical component of…
-
Prinz Eugen Ransomware Uses Go-Based Encryptor to Target Fresh Files and Evade Forensics
A customer compromised by a newly observed ransomware family we attribute to the Prinz Eugen group. The encryptor is a purpose-built Go binary that departs from many first-wave samples by combining deliberate file-targeting, modern cryptography, and anti”‘forensic measures traits that raise both technical and operational concerns for defenders. The intrusion likely began with compromised RDP…
-
Beats Studio Buds Vulnerability Lets Attackers Within Bluetooth Range Access Microphone
Apple has revealed a significant security vulnerability affecting Beats Studio Buds, which could allow attackers within Bluetooth range to access a device’s microphone without user consent. This issue, identified as CVE-2025-20701, was addressed in Beats Firmware Update 1B211, released on June 16, 2026. According to Apple’s advisory, the flaw impacts devices that are not yet…
-
Chinese Cyber Operations Shift From APT Groups to Composite Responsibility Model
Chinese state-linked cyber activity has moved decisively away from the neat, single-actor narratives that dominated early attribution toward an ecosystem model in which responsibility is distributed across military units, intelligence services, private firms, and criminal-style intermediaries. Official advisories characterized some companies as providers of cyber-related products and services to Chinese intelligence; the UK’s NCSC said…
-
FortiBleed Campaign Targets FortiGate Devices to Harvest VPN and Admin Credentials
Tags: advisory, attack, authentication, credentials, cyber, data-breach, exploit, fortinet, threat, vpnFortinet has issued a security warning about ongoing credential-harvesting attacks targeting FortiGate devices in a campaign known as “FortiBleed.” Threat actors are exploiting weak authentication practices rather than any newly disclosed vulnerabilities. A PSIRT advisory released on June 19, 2026, by Carl Windsor indicates that the attackers are reusing previously exposed credentials from earlier incidents,…
-
GitHub Actions Checkout Adds Protection Against Malicious pull_request_target Workflows
GitHub has implemented a major security enhancement in its Actions ecosystem with the release of actions/checkout v7, which aims to address a long-standing class of vulnerabilities known as “pwn requests.” This update was announced on June 18, 2026, and introduces safer defaults for workflows triggered by the pull_request_target event. This event is one of the…
-
Sapphire Sleet Hijacks npm Maintainer Account to Publish Poisoned Mastra Packages
A widespread npm supply”‘chain compromise to Sapphire Sleet, a North Korean state actor, after the takeover of an npm maintainer account enabled the mass publication of poisoned Mastra packages that silently delivered a multi”‘stage implant. The campaign, disclosed June 19, 2026, began when the attacker gained control of the ehindero maintainer identity an account with…
-
pgAdmin 4 Released with Patches for Seven Vulnerabilities and Feature Enhancements
pgAdmin 4 version 9.16 has been released by the pgAdmin Development Team, introducing significant security improvements along with feature enhancements and bug fixes. This update addresses seven vulnerabilities, tracked as CVE-2026-12044 through CVE-2026-12050, and includes 64 bug fixes and various usability upgrades. As one of the most widely used open-source management tools for PostgreSQL environments,…
-
Who pays when you gate cyber-capable AI models?
In this interview with Help Net Security, Jaya Baloo, COO CISO at Aisle, examines the debate over restricting access to cyber-capable AI models. She lays out the … First seen on helpnetsecurity.com Jump to article: www.helpnetsecurity.com/2026/06/22/jaya-baloo-aisle-gating-cyber-capable-ai-models/
-
OXLOADER Uses MBA Obfuscation and Control-Flow Flattening to Bypass Static Detection
A previously undocumented Windows loader, tracked as OXLOADER, that combines sophisticated obfuscation and unconventional staging to evade static detection and sandbox analysis while delivering the new CASTLESTEALER infostealer via malvertising. The campaign leveraged malicious Google Ads impersonating Node.js and API Monitor, redirecting victims through intermediary domains to Storj-hosted batch scripts that download and execute OXLOADER…
-
Massive GitHub Attack Injects Malware into 10,000 Compromised Repositories
A large-scale malware distribution campaign utilizing GitHub repositories has been uncovered. This coordinated effort weaponized over 10,000 repositories to deliver Trojanized payloads. The activity was first identified on June 18, 2026, and highlights significant gaps in automated detection and monitoring of repositories on one of the world’s most widely used developer platforms. Massive GitHub Attack…