GitHub stories
Security teams gain rollback and policy controls as autonomous Claude agents begin writing and deploying code at machine speed.
Security teams are under pressure as attacks can now begin before patches land, making early risk detection critical.
Teams risk wasted cycles and quality slips unless staff can judge when AI output fits the system and when it simply looks right.
Nearly 100 organisations were hit in a six-week phishing spree that used GitHub repositories and Visual Studio Code tools to infect developers.
CrowdStrike said state-backed espionage and extortion are surging as AI assets inside tech groups draw hackers seeking code, models and access.
Developers will gain on-device AI, coding agents and faster builds as Apple widens its software toolset across iPhone, Mac and Vision Pro.
Developers and enterprise customers will get more AI controls as Microsoft adds agents, in-house models and security tools across its software stack.
The compact desktop aims to cut cloud costs for AI developers by letting them fine-tune and run large models locally on Windows.
Marketplace bookings through Microsoft rose by double digits as New Relic deepened integrations aimed at helping customers manage AI-era software risk.
Developers can now manage multiple AI coding agents in one place as GitHub tests a desktop Copilot app with worktree automation and review tools.
Rising vulnerability volumes are outpacing fix times, prompting HackerOne to roll out an AI system that feeds confirmed threats into developer tools.
Security teams can now apply the same rules to AI-generated code across development and deployment, as Salt broadens its platform to curb flaws earlier.
More than half of patched flaws in major DevOps tools were high or critical in 2025, putting software supply chains at greater risk.
European ministries face a stealthier cyber-espionage campaign as Webworm shifts to Discord and Microsoft cloud tools to steal data.
The two-year pledge is aimed at easing early software costs for founders and tightening Zendesk's grip on startup buying decisions.
Developers using open-source tools face heightened supply-chain risk after the botnet lost all four of its command channels.
A Floxy study warns developers that Google's coding assistant keeps code for 540 days and defaults to training on user data.
Customer support teams could gain more control as the new toolkit cuts developer queues and ties training into community hubs.
Enterprises may be able to deploy governed multi-agent AI systems in days, as Kore.ai ties its Artemis platform to Microsoft Azure.
The new connectors let teams keep AI-generated work visible on a shared canvas, reducing siloed chats and manual rework across apps.
