Custom variants of OpenAI's ChatGPT promoted in sponsored Google results are directing unsuspecting users to malicious sites that use ClickFix attacks to deliver malware.

Meta says its Muse AI agent cannot access a user’s Messages without explicit permission, disputing a journalist’s account that the agent read his private messages while the required Mac setting was…

Some of the flaws could allow remote attackers to execute arbitrary code or escape the browser sandbox.

Shieldworkz analysis finds Adif web infrastructure served as entry point for Renfe compromise in AI-assisted cyber breach.

A nonprofit in California is doing what Hugging Face has not—attempting to hold OpenAI legally accountable for the actions of its agents.

It’s been more than a year since [Henrik Forstén] built the first iteration of his synthetic-aperture radar (SAR) imaging drone, and he’s certainly been productive in the meantime. Not only did he …

Security vetting tools and processes weren’t designed with obscure AI services in mind, according to TrendAI.

More than 80,000 proxy servers have been observed cloaking the origin of traffic to frontier models, likely using AI credentials harvested via information stealers, phishing campaigns, and…

NeedyMantis maintains long-term access in targeted intrusions, using DLL sideloading and HTTPS-to-WebSocket command-and-control.

AI agents need context-aware security beyond traditional models. Workflow data provides the governance guardrails that make AI safe and scalable.

Lunex uses BYOVD to disable kernel security callbacks before stealing browser credentials and cryptocurrency wallets.

Facial recognition is everywhere now. Soon it's going to be right in your face, too.

AI attackers move at machine speed. Security validation needs to follow every meaningful change.

A high-severity zero-day vulnerability affects the TDengine time-series database used across industrial, IoT, energy, and automotive environments.

Ox Security found security shortcomings in analysis of over 15,000 MCP servers

CISOs and CFOs must align on cybersecurity strategy to protect assets, manage risk, and enable business growth in today's threat landscape

Japanese car-sharing service Times Car has confirmed that approximately 6.6 million user accounts were compromised in a cyberattack disclosed late last week.

http-terminator is PortSwigger's AI pipeline for discovering HTTP request-smuggling bugs. Which stages run, its dependencies, and its testing limits.

Cyberattacks are exploiting trusted access points like compromised credentials and session cookies. Discover how organizations are gaining visibility beyond their perimeter to stay ahead of threats.

AI employees are raising concerns about an extinction-level event. But there’s a more immediate threat.

Human contractors are reviewing Copilot users’ prompts and uploaded images, according to internal documents obtained by 404 Media. The contractors are also bombarded with users’ requests for sexual…

The Medicaid IDs and other information of Medicaid and DC Healthcare Alliance beneficiaries were exposed.

Hackers stole personal data from a Polish healthcare software provider in the latest cyberattack to hit the country’s medical sector in recent months.

Google Ads deliver a tech support scam kit that shows fake security alerts, makes browsers appear locked and targets users across hundreds of organizations.

The Carbonato botnet compromises exposed Docker hosts, deploys an AI agent to steal AI API keys, and then spreads to other exposed Docker services.

Infostealer logs exposed AI account credentials and sessions tied to more than 80,000 corporate domains, creating risks ranging from stolen conversations to LLMjacking. SOCRadar examines the growing…

Researchers found more than 16,000 misconfigured Supabase databases exposing readable tables with personally identifiable information, passwords, or authentication tokens.

Apple patched CVE-2026-86950, a CoreGraphics flaw that may have been exploited in targeted attacks via maliciously crafted

Google support maps out how Googlebooks can take over where ChromeOS left off.

A quarter of victims of deepfake attacks have lost over $1m. CISOs worry that boardrooms don’t understand the threat

Muse agreed a lowball price and sent a Marketplace buyer to a user’s building, he says. Meta says Muse usually asks for permission.

Legislation introduced in the Senate on Wednesday would pause AI research until the creation of a federal department that would regulate the technology.

This series aims to set out what a meaningful, multistakeholder, and globally relevant AI accountability mechanism should look like.

NVIDIA has launched a platform pairing runtime controls with hardware monitoring for AI agents

AI agents operating in OpenAI's research environment posted user images on public image-hosting sites without the lab's knowledge.

OpenAI is facing a firestorm of criticism over incidents where its agents made unauthorized use of credentials, leaked data, and hacked websites.

Recent hacks have shown that the law is lagging when it comes to holding companies accountable.

ArsTechnica is reporting on a “new” attack against RSA, one that bypasses factoring. First, this attack isn’t new. The original research is from 2007. What is new is the implementation. Second, it is…

AI companies’ insistence that their technology is making breakthroughs, not just aiding scientists, is making real progress harder to recognize.

NCC Group reports that ransomware activity hits a 2026 high as the industrial sector bears 31% of attacks and Qilin dominates.

The extortion group has modified its exploit in new attacks targeting the PeopleSoft vulnerability CVE-2026-35273.

New WEF post observes satellite communications growth expands cybersecurity attack surface across IoT, utilities and critical infrastructure.

Earlier this month, Anthropic published a long report detailing all of the Claude misuses it detected. Daniel Meissler usefully summarized the report into 117 findings. A few of the highlights: AI…

Cloudflare has fixed a vulnerability in Containers and Sandboxes that allowed customers with a Workers Paid account to recover residual data from other customers' containers on the same physical host.

Major retailers were hit by sophisticated attacks performed on the cheap.

Security teams received urgent calls to disconnect servers before authorities confirmed critical zero-day flaws.

Keio Corporation (Keio), a major private railway operator in Japan, said its network was hit by a ransomware attack over the weekend, disrupting  some of its business systems.

God damn it, we're back in the room again. Yes, that sound in your ears is screaming. The footgun has gone off again, shockingly, and we are yet again dealing with a situation where the entire world…

Forcepoint research shows fake email threads can manipulate AI email summarizers without hidden text or direct instructions, inserting fabricated details.