AI-powered attacks, AI-powered defenses, and the vulnerabilities of the models themselves.
602 articles
WhatsApp has successfully identified and disrupted spear phishing attempts orchestrated by the spyware firm NSO. This action reinforces platform security against state-sponsored or commercial surveillance tools.
Meta Newsroom · 103d ago
Rubrik CEO Bipul Sinha highlights that the rise of AI agents significantly expands the attack surface for enterprises. He advocates for a strategic shift from traditional attack prevention to rapid data recovery.
Bloomberg Technology · 106d ago
Upwind has announced a new 'Security for AI' initiative, shifting its focus toward securing the entire AI stack rather than treating AI security as a standalone product. This strategy integrates security directly into agentic AI workflows and infrastructure.
The Next Web (TNW) · 107d ago
The article argues that AI's primary cybersecurity risk isn't just finding vulnerabilities, but the rapid weaponization of patched exploits and the scaling of sophisticated social engineering. It highlights that the automation of post-exploitation tasks poses a greater threat to digital infrastructure than initial vulnerability discovery.
LessWrong AI · 107d ago
Dashlane disclosed how attackers successfully downloaded encrypted password vaults by targeting a large volume of users. The company emphasizes that the vaults remained encrypted, limiting the immediate impact of the data exposure.
Ars Technica · 107d ago
CrowdStrike reported revenue figures in line with analyst expectations. The company highlights the growing urgency for cybersecurity solutions as hackers increasingly leverage powerful AI tools.
Bloomberg Technology · 108d ago
Dashlane issued a security advisory confirming that 20 encrypted user vaults were stolen. The company has provided limited details regarding the breach, maintaining silence on specific vectors.
Ars Technica · 108d ago
Palo Alto Networks raised its profit outlook, citing strong demand for security services driven by AI-related threats. Companies are increasingly investing in AI-powered security to mitigate emerging risks.
Bloomberg Technology · 109d ago
Investors are closely watching upcoming earnings reports from Palo Alto Networks and CrowdStrike to determine if the 37% rally in cybersecurity stocks is sustainable. These companies are central to the AI-driven cybersecurity landscape.
Bloomberg Technology · 109d ago
Anthropic is expanding its 'Project Glasswing' cybersecurity initiative by granting access to the Claude Mythos Preview model to approximately 150 additional organizations. This move aims to bolster security research and application development using their specialized AI model.
ITmedia AI+ (日本) · 109d ago
Smart App Control (SAC) is a Windows 11 security feature that uses AI and cloud intelligence to block untrusted or malicious applications. It functions as an automated gatekeeper, verifying digital signatures and analyzing behavior to prevent security threats in real-time.
Computerworld · 110d ago
The newly detailed FROST technique allows websites to measure SSD activity via browser-based JavaScript. This poses significant privacy risks by enabling side-channel attacks through storage access patterns.
Wired · 110d ago
The rapid advancement of quantum computing is creating significant vulnerabilities for current enterprise security standards. Organizations must prepare for a shift in cryptographic requirements to defend against future quantum-based threats.
ZDNet AI · 111d ago
Microsoft is threatening criminal action against a researcher known as Nightmare Eclipse for disclosing zero-day exploits without proper coordination. The company has also suspended the researcher's access to key security and development platforms.
The Verge · 112d ago
Microsoft is facing backlash after threatening legal action against a researcher who publicly disclosed unpatched vulnerabilities in Windows Defender and BitLocker. The cybersecurity community views this as an aggressive move against responsible disclosure practices.
The Next Web (TNW) · 112d ago
Google has enabled Device Bound Session Credentials (DBSC) by default in Chrome on Windows. This security feature binds session cookies to the physical device, rendering stolen cookies useless to attackers.
Digital Trends · 113d ago
Researchers have identified a side-channel attack that tracks user browsing habits by monitoring SSD timing characteristics. This method bypasses traditional tracking mechanisms like cookies or fingerprinting.
cnBeta (Full RSS) · 114d ago
California Attorney General Rob Bonta has filed a lawsuit against 23andMe, alleging the company failed to protect sensitive genetic data and misled the public regarding the severity of a 2023 breach. The legal action highlights growing regulatory scrutiny over how biotech firms handle massive datasets used for AI-driven health research.
BBC Technology · 114d ago
NATO has signed non-commercial cybersecurity agreements with Microsoft, Palo Alto Networks, and ESET. These partnerships aim to strengthen collective cyber resilience across the alliance.
The Next Web (TNW) · 115d ago
This article explores whether using cutting-edge AI models like Anthropic's Claude Mythos is sufficient for robust cybersecurity, using Mozilla's Firefox development as a case study.
ITmedia AI+ (日本) · 115d ago
OpenAI is implementing new measures to support election integrity ahead of global 2026 elections. These initiatives focus on improving information access, assisting cyber defenders, and enhancing AI transparency.
OpenAI News · 116d ago
The European Central Bank is mandating that financial institutions strengthen their IT systems against new vulnerabilities introduced by AI models. This follows a high-level meeting focused on the evolving threat landscape posed by advanced artificial intelligence.
Bloomberg Technology · 118d ago
The surge in AI-generated code and the emergence of advanced models like Anthropic's Mythos have created new security vulnerabilities. This has led to a significant increase in demand for specialized cybersecurity engineers.
New York Times Technology · 118d ago
New research reveals that 62% of senior executives knowingly use unapproved 'shadow AI' tools, prioritizing productivity over security compliance. This significantly outpaces the 31% adoption rate among general employees, creating massive enterprise security vulnerabilities.
TechRadar AI · 120d ago
Flipper Devices is developing the Flipper One, a portable Linux-based multi-tool. It features high-end hardware including an 8-core processor and extensive connectivity options.
The Verge · 121d ago
Industry practitioners suggest that concerns regarding the potential for Anthropic's Mythos AI model to facilitate unfettered hacking are largely exaggerated. Current analysis indicates that the actual security risks are measured and manageable.
iTNews Australia · 122d ago
New details reveal that Russian hackers involved in the 2020 SolarWinds breach gained deep access to U.S. Treasury Department emails.
Bloomberg Technology · 122d ago
Socket, a cybersecurity startup focused on protecting open-source code from nation-state hackers, has reached unicorn status. The company provides tools to secure software supply chains, which is increasingly critical for AI-driven development environments.
Bloomberg Technology · 122d ago
Verizon's latest Data Breach Investigations Report (DBIR) reveals that attackers are shifting focus to SMS and voice calls. This trend is driven by the increasing effectiveness of modern email security defenses.
ZDNet AI · 122d ago
A previously undisclosed 0-day vulnerability in Huawei enterprise routers has been identified as the cause of a major nationwide communication outage in Luxembourg last year. The incident disrupted mobile, landline, and emergency services for over three hours.
cnBeta (Full RSS) · 122d ago