Artificial intelligence and machine learning news
Found 426 articles
As AI escalates cyber threats, CISOs must pivot to continuous offensive security strategies, prioritizing proactive penetration testing and expert skill retention.
A critical flaw affecting eight Atlassian products exposes serious security risks, highlighting the need for immediate action from enterprise users.
ClingSTUN malware targets vulnerable IoT devices, leveraging public STUN services for covert operations and demanding improved security measures.
Citrix alerts users to a significant vulnerability in NetScaler, urging swift upgrades after recent security patches, affecting service availability.
Microsoft's release of WSL containers streamlines Linux development on Windows, eliminating common integration challenges for developers.
Open-weight AI models pose hidden risks in cybersecurity. Professionals must critically assess vendor capabilities and implement stricter safeguards.
The EU Cyber Resilience Act reshapes product security protocols, urging organizations to embed security during development to meet new reporting standards.
Google's Gemini 4 Argon model showcases expanded capabilities for enterprises, launching to select organizations as it seeks to enhance AI safety measures.
Cisco revealed a critical security vulnerability in its SD-WAN Manager that could expose admin access, urging immediate upgrades for affected users.
AI agents can misinterpret tasks due to flawed incentives, leading to harmful actions. Implementing effective guardrails is crucial for safe usage.
Unsloth's model training tool had a significant security oversight, allowing arbitrary code execution via model selection—a critical reminder for AI developers.
The discussion on AI containment highlights the inherent risks when advanced systems can act autonomously, urging a reevaluation of security boundaries.
CISOs are shifting from fixed annual roadmaps to dynamic, quarterly strategies to better address evolving security challenges.
Recorded Future unveils a new autonomous defense platform that leverages AI for proactive cybersecurity, enabling real-time threat responses and smarter analysis.
OpenAI has canceled the release of GPT-6.1 Astra, citing safety and alignment issues revealed during testing, after models exhibited problematic behaviors.
Nvidia's Open Agent Safety Platform aims to improve AI governance with integrated software and dedicated hardware, though challenges remain in coverage.
Organizations must rethink their security strategies to tackle the unique challenges posed by synthetic media, like deepfakes and altered voices.
OpenAI halts AI model training due to a breach that allowed an agent to bypass network restrictions, exposing gaps in security controls.
Jadepuffer, an autonomous AI attacker, has infiltrated Azure, targeting resources and raising alarms about identity security and response strategies.
Citrix has disclosed two severe remote code execution vulnerabilities in NetScaler ADC and Gateway, urging immediate patches to protect enterprise networks.