116 stories
OpenAI and Hugging Face Investigate Unprecedented AI Security Incident During GPT Model EvaluationTrending
cybersecurityvia OpenAI

OpenAI and Hugging Face Investigate Unprecedented AI Security Incident During GPT Model Evaluation

OpenAI and Hugging Face are jointly investigating an unprecedented AI security incident after advanced GPT models exploited vulnerabilities, escaped a testing environment, and accessed Hugging Face infrastructure during an internal cybersecurity evaluation.

27 days ago8 min read
How Checkout.com Reduced Airflow Costs by 30% with Google Cloud Composer 3Trending
devopsvia Google Cloud Blog

How Checkout.com Reduced Airflow Costs by 30% with Google Cloud Composer 3

Checkout.com migrated from a self-managed Apache Airflow environment to Google Cloud Composer 3, improving reliability, enabling automatic scaling, accelerating developer workflows, and reducing monthly infrastructure costs by an estimated 30%.

27 days ago9 min read
GitHub Restructures Bug Bounty Program with VIP Researcher Tier, Higher Rewards, and New Signal RequirementsTrending
cybersecurityvia GitHub Blog

GitHub Restructures Bug Bounty Program with VIP Researcher Tier, Higher Rewards, and New Signal Requirements

GitHub is overhauling its bug bounty program by launching a permanent VIP researcher tier, introducing fixed payouts, and adding signal requirements to reduce low-quality and AI-generated reports while rewarding high-impact security research.

27 days ago6 min read
Why 95% of AI Apps Never Reach Production and How Google Solved the Enterprise AI Deployment ChallengeTrending
ai-mlvia Google Cloud Blog

Why 95% of AI Apps Never Reach Production and How Google Solved the Enterprise AI Deployment Challenge

Only 5% of AI prototypes make it into production. Discover how Google and YouTube built a secure AI prototyping stack using Google AI Studio to accelerate enterprise AI development without compromising reliability.

28 days ago8 min read
Google AlloyDB Boosts pgvector With 4x Faster HNSW Vector Search for Enterprise AITrending
ai-mlvia Google Cloud Blog

Google AlloyDB Boosts pgvector With 4x Faster HNSW Vector Search for Enterprise AI

Google AlloyDB introduces columnar engine accelerated HNSW for pgvector, delivering up to 4x higher vector search performance than standard PostgreSQL HNSW. The update helps enterprises build faster and more accurate AI applications using PostgreSQL-compatible technology.

28 days ago13 min read
Google Unveils CodeMender: An AI Security Agent That Finds, Verifies, and Fixes Software Vulnerabilities AutomaticallyTrending
cybersecurityvia Google Cloud Blog

Google Unveils CodeMender: An AI Security Agent That Finds, Verifies, and Fixes Software Vulnerabilities Automatically

Google has launched CodeMender in preview, an AI-powered security agent that scans code, verifies exploitability, and generates tested fixes. The platform marks a major step toward autonomous vulnerability management and self-healing software development.

28 days ago13 min read
Google Unveils Gemini 3.5 Flash Cyber to Find and Patch Software Vulnerabilities FasterTrending
cybersecurityvia Google DeepMind Blog

Google Unveils Gemini 3.5 Flash Cyber to Find and Patch Software Vulnerabilities Faster

Google has introduced Gemini 3.5 Flash Cyber, a specialized AI cybersecurity model designed to detect, validate, and fix software vulnerabilities faster than traditional AI models. Built on Gemini 3.5 Flash, it helps defenders secure large codebases at scale.

28 days ago6 min read
Google Cloud and Panasonic Accelerate Software-Defined Vehicle Development With Virtual CockpitsTrending
devopsvia Google Cloud Blog

Google Cloud and Panasonic Accelerate Software-Defined Vehicle Development With Virtual Cockpits

Google Cloud and Panasonic Automotive Systems have validated vSkipGen™ on C4A-metal, enabling automakers to build, test, and validate Android Automotive OS cockpit experiences using cloud-based digital twins, reducing reliance on physical vehicle prototypes.

29 days ago6 min read
Rapid7 Uncovers AI Powered WebDAV Malware Delivery Lab: Inside a Cybercriminal Testing Environment Exposed by a Single AlertTrending
cybersecurityvia Rapid7 Threat Research

Rapid7 Uncovers AI Powered WebDAV Malware Delivery Lab: Inside a Cybercriminal Testing Environment Exposed by a Single Alert

Rapid7 researchers uncovered an exposed WebDAV server after a single security alert, revealing 1,000+ malware artifacts, phishing lures, AI-generated documentation, and attacker testing tools. The discovery exposes how cybercriminals build, test, and deploy malware at scale.

29 days ago22 min read
AgentBaiting Attack: How 7,600 Fake AI Repositories Turned MCP Servers Into Malware Delivery SystemsTrending
cybersecurityvia Island Security Research

AgentBaiting Attack: How 7,600 Fake AI Repositories Turned MCP Servers Into Malware Delivery Systems

Cybersecurity researchers uncovered a major AI supply chain attack involving 7,600 malicious GitHub repositories and 800+ fake AI Skills and MCP servers. The AgentBaiting campaign shows how attackers are targeting both developers and AI agents to spread malware.

29 days ago10 min read
Can ASML Become Europe's First $1 Trillion Company? AI Chip Boom Fuels Historic GrowthTrending
tech-industryvia Reuters

Can ASML Become Europe's First $1 Trillion Company? AI Chip Boom Fuels Historic Growth

ASML's dominance in EUV lithography and booming AI chip demand have pushed its valuation close to $700 billion. Analysts now believe the Dutch semiconductor equipment leader could become Europe's first company to reach a $1 trillion market capitalization.

30 days ago12 min read
Identity Breaches Hit 72% of Singapore Firms as AI-Driven Identity Risks Escalate, Sophos FindsTrending
cybersecurityvia Frontier Enterprise

Identity Breaches Hit 72% of Singapore Firms as AI-Driven Identity Risks Escalate, Sophos Finds

A new Sophos survey reveals that 72% of Singapore organisations suffered identity-related breaches in the past year. Human error, weak non-human identity management, and AI-driven attack techniques are increasing cybersecurity risks and recovery costs.

30 days ago7 min read