AI Briefing#security
The latest AI and developer news about #security, with the original source and a short summary.
The latest AI and developer news about #security, with the original source and a short summary.
105M fuzzing runs uncovered a Lean runtime heap overflow and a lean-zip DoS.
An LLM security benchmark for finding known vulnerabilities in real codebases
Over 30 WordPress plugins were infected with backdoors after being acquired.
This is a case of an attack on open source developers that targets not vulnerabilities but trust.
After iOS 26.4, a Czech háček input error blocked device unlock.
Structural loopholes were revealed in 8 AI agent benchmarks.
Microsoft suspended accounts of major open-source developers without prior notice.
The download links for CPU-Z and HWMonitor were tampered with, distributing malware.
An experiment log scraping and analyzing roughly 80,000+ Firefox extensions.
After resolving a Microsoft signing issue, WireGuard has put out a new release for Windows.
It points out the security risk that `npx -y ...@latest` automatically runs new npm code on every session, and suggests switching to a pinned-version local execution instead.
Cred, an open-source credential delegation middleware that solves OAuth token security issues for AI agents, has been launched.
Origin, an enterprise-grade AI Development Environment (CDE) with enhanced security and privacy, has launched.
Introduced automation methods using tools such as Certbot, acme.sh, and cert-manager in response to the trend of validity periods being shortened to 200 days or less.
A supply chain attack was discovered in the npm **axios** package, requiring security measures for users of certain versions.
Anthropic has launched the Compliance API, which lets admins programmatically audit activity logs on the Claude Platform.
It guides how to manage code vulnerabilities and leaked secrets by using GitHub's security tools.
AI21 Labs has successfully completed its annual SOC 2 audit to comply with data protection and security standards.
Meta automates large-scale security migrations with a secure-by-default Android framework and generative AI.
A developer's journey of growth from DB and security to AI, built on reasonable choices rather than right answers.
Starting June 30, 2026, Vercel will discontinue support for the legacy DHE cipher suite.
Notion Workers leverages Vercel Sandbox's microVMs to safely run untrusted code at scale in isolated environments.
The core themes of AWS re:Invent 2025 were Agentic AI, security, and EKS management automation.
Yeogieotae's DevOps team built its own sensitive information management platform, **Secrethub**, to strengthen security and improve operational efficiency.