Docker stories
Open-source maintainers can now spot stolen build secrets sooner, as Tracebit removes repository limits for public projects and widens free coverage.
Supply chain compromises helped actively exploited flaws to 215 in the first half of 2026, as attackers used trusted tools and platforms to steal credentials.
Teams testing software integrations could soon avoid live APIs and shared sandboxes, as LocalStack adds WonderTwin AI's application emulators.
With Fortune 500 firms expected to deploy vast armies of agents, the alliance aims to curb access risks as AI tools spread across business systems.
Enterprises could cut AI ingestion errors and token costs as the new tool keeps sensitive documents on premises or in a chosen cloud.
The experimental setup lets AI teams run untrusted model code in isolated Ray clusters on Google Kubernetes Engine without exposing host containers.
Malicious packages are now spreading faster across code repositories, with Google saying open source ecosystems face the sharpest rise in risk.
Security and compliance teams can now patch buildpack-based containers from a single hardened base, rather than chasing Dockerfiles across repositories.
The platform lets security teams run AI pentests without exposing customer infrastructure data to external models.
Windows fleets could be left blind to malware once attackers gain admin rights, as Bitdefender says bind links can fool endpoint security tools.
UK and Ireland partners will gain wider access to Docker's container platform as Exclusive Networks expands its cloud-native software portfolio.
Users of Dify's cloud service could have had private chats and files exposed after Zafran Security disclosed four flaws in the AI platform.
The expansion gives IT teams central control over AI agent permissions, reducing risky static keys and easing reviews as workplace use widens.
Older Liquibase Community users can now check release-by-release vulnerabilities in a free public library covering Docker images and binaries.
Rising demand for secure AI software development has prompted Sonatype to expand its leadership team and scale operations globally.
Developers can now run OpenClaw and Hermes without setting up servers, as Cloudways adds managed hosting to cut deployment and maintenance work.
Banks and funds can now build AI-enabled workflows on one governed system, as 3forge aims to reduce integration work and oversight risks.
U.S. agencies can now train and keep control of AI models on isolated systems, with Palantir and NVIDIA targeting sensitive government work.
Production users can now route generative AI requests through a stable open source gateway, with Bloomberg already running it and Nutanix adopting it.
A free account could have let attackers alter Zapier-maintained packages and hijack logged-in users' browser sessions, researchers said.
