Security stories
The open standard aims to let organisations verify AI agents before they are trusted to act across networks, systems and services.
Security teams can now spot suspicious repository activity and restore only malicious changes in GitHub and Azure DevOps after identity-based attacks.
The move puts autonomous AI agents and GPU clusters under open source scrutiny as NVIDIA seeks to shape the Kubernetes stack around them.
NetSuite partner Annexa's Matthew Owens said the NetSuite Next upgrade is nearly effortless, but finance is slower putting AI agents to work.
The fresh cash will expand Denmark's Gefion supercomputer capacity for public bodies, researchers and regulated industries as AI demand grows.
The deal aims to cut the operational sprawl facing companies as AI agents push analytics, transactions, monitoring and security onto one platform.
Customers can now use natural language prompts to diagnose issues and assess upgrade risks on the updated hybrid cloud platform.
Investors in decentralised finance can now see S&P Global Ratings' vault risk analysis directly onchain as lending deposits near USD $10 billion.
Large enterprises are spending more than 13,900 engineering hours a year fixing AI-assisted code as governance costs spill into security and compliance budgets.
One in three large Asia Pacific enterprises will struggle to govern shadow AI agents as staff-built tools outpace internal controls, Forrester says.
Access to Perplexity's new Teammates feature is limited to individual Max subscribers, leaving Pro and enterprise users waiting for a broader rollout.
Rapid revenue growth and a bigger AI pipeline have helped KloudStax secure elite Google Cloud status, widening customer access to technical support.
The update aims to help businesses run AI and traditional systems together as security concerns and operational complexity rise across hybrid estates.
Partners can now trial private AI workloads locally as the hosted kit helps them assess compliance, latency and sovereignty before production.
The move centralises Verint's AI product development as the software group tries to turn customer demand for measurable results into faster delivery.
OS Bondi 3.0 to cut due diligence time and give deal teams faster risk flags, while keeping sensitive transaction data within a governed secure environment.
Yet most Engineering and Technology leaders are still relying on old metrics, despite 96% saying AI is improving software development and delivery.
Demand is building for agent-ready hardware, with Otto saying it has presold 12,500 of its $599 screenless computer in five months.
Fewer than half of Australian firms say their AI foundations are ready for more autonomous systems, Cisco's research found.
Only 12 of 50 sites were fully ready, raising the risk Australian organisations lose sales as AI agents take over online tasks.