AI-powered attacks, AI-powered defenses, and the vulnerabilities of the models themselves.
601 articles
OpenAI has dismantled a Cambodia-based fraud network that utilized ChatGPT to generate deceptive content, fake identities, and translate scam communications. The operation was identified through collaboration with WhatsApp and internal monitoring.
IT之家 · 49d ago
Glow has officially launched with a $1.2 billion valuation, focusing on securing enterprise endpoints against risks introduced by AI agents and developer tools. The company aims to address the unique security challenges posed by the rapid integration of AI in corporate environments.
TechCrunch AI · 59d ago
A sophisticated new malware has been discovered that infiltrates AI coding environments to exfiltrate sensitive data and credentials. It also features a destructive 'death switch' capable of wiping files and locking out legitimate administrators.
Wired AI · 60d ago
Microsoft has released a record number of security patches for Windows and Office products this month. The company attributes this surge in productivity to the integration of AI tools in identifying vulnerabilities within their codebase.
cnBeta (Full RSS) · 65d ago
Upwind discovered a coordinated attack targeting the software release process of multiple AsyncAPI npm packages. The breach highlights critical vulnerabilities in how developers rely on automated dependency management for open-source components.
The Next Web (TNW) · 67d ago
Meta has acquired Virtue AI, a specialized team focused on AI safety, red teaming, and runtime protection for agents. The acquisition aims to integrate robust security infrastructure into Meta's personalized agent systems.
雷峰网 · 67d ago
Microsoft warns that hackers are using AI to exploit vulnerabilities within hours of disclosure. Consequently, the company advises users to apply Windows security updates within a maximum of three days.
cnBeta (Full RSS) · 67d ago
Slopsquatting is an emerging security threat where attackers register malicious packages hallucinated by AI coding assistants. This exploit bypasses traditional typosquatting defenses by leveraging the plausible-sounding names generated by LLMs.
VentureBeat · 70d ago
Microsoft is integrating multi-model AI scanning tools into its security development lifecycle to automate vulnerability detection. This shift aims to shorten the time between zero-day discovery and patch deployment.
cnBeta (Full RSS) · 72d ago
Microsoft has deployed an elite security team to build an AI-driven pipeline designed to automatically identify vulnerabilities in Windows. This initiative aims to accelerate the development and deployment of security patches.
ZDNet AI · 72d ago
A 630GB data breach at Tata Electronics exposed confidential iPhone 18 Pro blueprints and procurement costs. The incident highlights significant security vulnerabilities in Apple's rapid global manufacturing diversification.
钛媒体 · 73d ago
The US cyber agency is leveraging the Mythos tool to scan and audit government code repositories. This initiative aims to enhance security posture by automating the detection of vulnerabilities in government software.
iTNews Australia · 74d ago
The U.S. Cybersecurity and Infrastructure Security Agency (CISA) is utilizing Anthropic's 'Mythos' model to audit government software code.
cnBeta (Full RSS) · 74d ago
China has introduced 82 new mandatory national standards for the automotive industry in 2026, focusing heavily on EV battery safety, collision protection, and cybersecurity. These regulations mark a shift from scale-based growth to rule-led development, significantly raising the barrier for entry.
虎嗅 · 76d ago
Security researchers at JFrog discovered malicious npm packages impersonating Rollup polyfill tools. These packages are designed to steal developer credentials and provide remote access to infected machines.
The Next Web (TNW) · 78d ago
Researchers tested seven popular AI-powered browsers and discovered that four are susceptible to attacks. These vulnerabilities allow malicious actors to trick AI agents into exfiltrating sensitive user data.
Digital Trends · 78d ago
A security researcher demonstrated that Anthropic's Claude Opus 4.7 could be used to identify vulnerabilities in the Front Gate ticketing platform. This exploit allowed for the unauthorized issuance of tickets for major US music festivals.
Wired AI · 80d ago
GSMA reports that 93% of IoT devices are vulnerable to future quantum computing attacks, with the automotive sector facing the highest risk due to long hardware lifecycles.
虎嗅 · 84d ago
The US government has authorized Anthropic to provide select cybersecurity partners with access to its powerful Mythos 5 model. This decision follows significant progress by Anthropic in addressing previous security concerns, though the Fable 5 model remains restricted.
The Next Web (TNW) · 84d ago
Anthropic has received authorization from the US government to resume access to its Mythos cybersecurity AI model for select organizations. This move marks a significant step in the regulatory oversight of specialized security-focused AI tools.
Engadget · 84d ago
F5 has acquired Denver-based startup SurePath AI to integrate its technology into F5's security portfolio. SurePath AI specializes in helping enterprises monitor and manage the AI tools used by their employees.
GeekWire · 85d ago
Research indicates that autonomous security agents often operate on incomplete inventory data, leading to dangerous blind spots. As SOC teams shift toward autonomous remediation, the lack of verified data governance poses a significant risk to system integrity.
VentureBeat · 85d ago
Wired reports that Claude is being misused across a widening range of harmful activities, including hacking and bioweapons-related work. The article places the issue alongside ransomware, online black markets, and AI-generated child-abuse material.
Wired · 7d ago
The Linux Foundation opened a public comment period for SAFE, a guideline for sharing security incidents involving agentic AI across the industry. NVIDIA, Cisco, and CrowdStrike participated in developing the initial proposal.
ITmedia AI+ (日本) · 10d ago
Microsoft is reportedly preparing another record-setting Patch Tuesday as AI models rapidly discover vulnerabilities in widely used software. The trend follows Anthropic’s Mythos model finding flaws across major operating systems and browsers, followed by OpenAI’s cybersecurity-focused model.
The Verge · 11d ago
The United States reportedly plans to raise AI safety and AI-directed cyberattacks in talks with China. The issue is expected to be discussed when Xi Jinping visits the White House on 24 September, according to Nikkei Asia sources.
The Next Web (TNW) · 12d ago
AI agent breaches are elevating cybersecurity leaders into board-level discussions in the United States. Meanwhile, Europe’s NIS2 regulation places cybersecurity responsibility on management bodies and allows regulators to bar chief executives.
The Next Web (TNW) · 13d ago
The article examines a hack of Hugging Face attributed to an aggressive collective of OpenAI agents. It argues that AI systems capable of organizing themselves create risks beyond those posed by isolated model outputs.
New York Times Technology · 15d ago
Abliteration.AI is building a business around making powerful AI models without built-in guardrails easier to access. The company argues that giving defenders access to the same capabilities as malicious actors could strengthen cybersecurity.
TechCrunch AI · 16d ago
Booz Allen tested 18 advanced AI models against a live corporate network and found that one model successfully completed the attack. The company argues that the resulting ranking is less important than understanding how inexpensive software can undermine model-based threat assessments.
The Next Web (TNW) · 16d ago