A cybersecurity researcher has shared a PoC for a Windows flaw called “MiniPlasma.” This lets attackers get SYSTEM privileges on fully updated Windows systems. The exploit was shared by a researcher called Chaotic Eclipse, or Nightmare Eclipse. They released the source code and a working file on GitHub after saying …
Read More »PoC released
Microsoft’s MDASH VS Anthropic’s Mythos VS OpenAI’s Daybreak
The newly launched of Microsoft MDASH, OpenAI Daybreak, and Anthropic Mythos shows a big change in cybersecurity. These tools go beyond basic chat functions by using groups of special models to find, discuss, and fix serious problems in important systems. MDASH and Daybreak aim to make security stronger during development, …
Read More »Microsoft’s new multi-model MDASH AI System discovered 16 Windows Flaws
Microsoft has unveiled a new AI-based system called MDASH to help find and fix problems on a large scale. They also said some customers are testing it in a private preview. MDASH is designed as a model-agnostic system that uses bespoke AI agents for different vulnerability classes to autonomously discover, …
Read More »OpenAI Unveils “Daybreak” To Automates Vulnerability Detection and Fixing
OpenAI unveils Daybreak, a frontier artificial intelligence system built specifically for cyber defenders. Daybreak aims to change how developers find, deal with, and fix serious weaknesses in software. By focusing on being ready ahead of time instead of just fixing problems after they happen, Daybreak helps stop bad actors before they …
Read More »Hackers misuse Claude.ai and Google ads chats to spread Mac malware
As AI is advancing unpredictably cyber criminals also change their attack pattern which make challenge for cyber resilience. Attackers are abusing Google Ads and legitimate Claude.ai shared chats in an active advertising campaign. Users looking for “Claude mac download” might find ads that show claude.ai as the site to visit. …
Read More »DDoS Attack Makes 2.45 Billion Requests Using 1.2 Million IP Addresses
A distributed denial-of-service attack hit a big user-content platform, sending an amazing 2.45 billion malicious requests in only 5 hours. Security company DataDome stopped the attack right away, so real users face no disruption. Experts looking into the event found that the attack used 1.2 million different IP addresses. The …
Read More »GPT-5.5 matches Claude Mythos in cyber attack tests: Report
Key Points: The UK’s AI Security Institute (AISI) tested OpenAI’s GPT-5.5 and found it can perform cyberattacks like Anthropic’s Claude Mythos Preview. GPT-5.5 is the second model, after Mythos, to fully complete a complicated enterprise attack test. This was done on a network without any active defenses. AISI sees this …
Read More »VECT 2.0 Ransomware Destroys Files On Windows, Linux & ESXi
Threat hunters say that the cybercrime group called VECT 2.0 is more like a wiper than ransomware. This is because of a big mistake in how it encrypts files on Windows, Linux, and ESXi systems, making recovery impossible even for the criminals. Check Point Research (CPR) looked at VECT 2.0 …
Read More »Claude-powered AI coding agent deletes entire company production database in 9 seconds
Jer Crane, the founder of PocketOS, explained how an AI agent delete his company’s production database. The agent, called Claude Opus 4.6 from Anthropic, used a regular access token to wipe out the production database and its backups on a platform called Railway. The founder posted his dilemma on a …
Read More »Censys Warns, 6 Million FTP Servers Still Exposed in 2026
A recent April 2026 report from security expert Himaja Motheram at Censys says that nearly 6 million internet-connected hosts still use the File Transfer Protocol (FTP). This shows a big 40% drop from the 10.1 million servers seen in 2024. This old protocol still brings a risk because many users keep …
Read More »
InfoSecBulletin Cybersecurity for mankind