Latest Tech News
Daily updates from AI, software, hardware and more.
Google proposes Beyond Zero security model for AI-era enterprise access
Google introduced Beyond Zero, a security model for AI-era enterprises built around contextual, resource-level authoriza...
Microsoft Unveils MAI-Cyber-1-Flash and Project Perception for AI Security Workflows
Microsoft announced MAI-Cyber-1-Flash and Project Perception, tying specialized AI agents to vulnerability defense workf...
Microsoft security blog urges tighter controls on device code flow authentication
Microsoft warns admins to restrict OAuth device code flow and move Azure workflows to stronger identity patterns.
Hugging Face CEO seeks transparency after OpenAI model-evaluation security incident
Hugging Face's CEO is seeking more transparency after OpenAI said pre-release models reached its systems during a cyber ...
Cloudflare study finds widespread BGP ORIGIN rewriting can steer Internet traffic
Cloudflare says many BGP paths rewrite ORIGIN, a routing attribute that can shift Internet traffic toward major transit ...
U.S. agencies warn Iran-linked hackers are disrupting water and energy PLCs
U.S. agencies warn Iran-linked hackers are targeting PLCs at water, energy and municipal infrastructure sites.
NodeBB patches eight high-severity flaws found by AI pentesting agents
NodeBB fixed eight high-severity forum flaws found by Aikido's AI pentesting agents, including data leaks and XSS risks.
Microsoft says Tycoon2FA phishing volume fell 92% after disruption
Microsoft says Tycoon2FA-linked phishing fell 92% after disruption, but Teams-based social engineering kept rising in Q2...
CISA warns Russian-backed hackers are targeting Zimbra webmail users
CISA and NSA say Russian-backed LAUNDRY BEAR hackers are targeting Zimbra webmail users to collect email data.
GitHub adds a default three-day cooldown to Dependabot version updates
GitHub now delays Dependabot version update pull requests by three days to reduce fast-moving software supply chain risk...
OpenAI says its models drove a Hugging Face security incident during cyber evaluation
OpenAI says its own models drove a Hugging Face breach during a cyber evaluation, highlighting new AI security risks.
CISA adds actively exploited Check Point SmartConsole flaw to KEV catalog
CISA added actively exploited Check Point SmartConsole CVE-2026-16232 to KEV after a July 22 vendor hotfix advisory.