Zero Networks extends microsegmentation to govern AI agents and block AI-driven attacks.
www.networkworld.com/article/4161...

Cloudflare wants to rebuild the network for the age of AI agents A slew of updates from Cloudflare includes its Mesh private networking fabric for AI agents and its Registrar API interface, which allows domain registrations to be handled by AI agents.

A slew of updates from Cloudflare includes its Mesh private networking fabric for AI agents and its Registrar API interface, which allows domain registrations to be handled by AI agents. www.networkworld.com/article/4160...

How to improve the SOC analyst experience -- and why it matters | TechTarget Learn how enterprise CISOs can improve the SOC analyst experience, or AX, to improve security outcomes, strengthen cyber resilience and retain key talent.

www.techtarget.com/searchsecuri...

Burned-out security analysts miss threats, take longer to investigate incidents and are more likely to quit. Here's how CISOs can improve the SOC analyst experience.

Databricks tested a stronger model against its multi-step agent on hybrid queries. The stronger model still lost by 21%. Databricks research tested a stronger foundation model against its multi-step Supervisor Agent on hybrid data queries spanning SQL and unstructured docs. The model lost by up to 38%, pointing to an architecture problem, not a model quality problem.

venturebeat.com/data/databri...

"RAG works, but it doesn't scale..If you want to make your agent even better, and you want to understand why you have declining sales, now you have to help the agent see the tables and look at the sales data. Your RAG pipeline will become incompetent at that task."

TIA Sets New Infrastructure Standards for the AI Data Center Era

“In an everything-connected world, TIA is trying to take a central focus on infrastructure,” TIA CEO David Stehlin told me.

www.datacenterknowledge.com/infrastructu...

Linux 7.0 debuts with some big changes for networking Why Linux 7's quiet TCP change could be the best thing to happen to your network.

Why Linux 7's quiet TCP change could be the best thing to happen to your network.
www.networkworld.com/article/4158...

Six weeks ago, Claude users reported degraded model capability. An Anthropic marketing replied saying users were imagining things.

Today, confronted with evidence, Anthropic says it did make changes, to "lowers total cost for users across the request mix."

These AI companies cannot be relied upon.

Threat research from Lumen Technologies focuses on telemetry from the internet backbone, where the largest ISPs interconnect, to reveal attack patterns that traditional endpoint security deployments cannot surface. www.networkworld.com/article/4156...

Aria Networks raises $125M and debuts its approach for AI-optimized networks Startup Aria Networks launches Deep Networking platform for AI factories, targets token efficiency.

Startup Aria Networks launches Deep Networking platform for AI factories, targets token efficiency. www.networkworld.com/article/4155...

Amazon S3 Files gives AI agents a native file system workspace, ending the object-file split that breaks multi-agent pipelines Amazon S3 Files mounts any S3 bucket directly into an agent's local environment, giving AI agents native file system access to enterprise data without a separate storage layer. AWS says the architecture eliminates the session state loss and shared-state failures that have blocked multi-agent pipelines in production.

venturebeat.com/data/amazon-...

Amazon S3 Files gives AI agents direct access to enterprise data. Analysts call it the missing link for agentic AI

OpenStack Gazpacho is a dish best served cold for hot cloud networks Open-source cloud infrastructure technology adds new compute and networking capabilities.

Open-source @openstack.org cloud infrastructure technology adds new compute and networking capabilities. www.networkworld.com/article/4153...

Inside the Push to Bring DC Power to Data Centers Two European alliances are collaborating to advance open DC distribution standards, aiming to improve data center energy efficiency.

Inside the Push to Bring DC Power to Data Centers - www.datacenterknowledge.com/energy-power...

How Lumen is dismantling decades of network complexity Lumen Technologies created a unified data model and deployed AI agents to transform overlapping legacy networks and accelerate service delivery.

Lumen Technologies created a unified data model and deployed AI agents to transform overlapping legacy networks and accelerate service delivery. www.networkworld.com/article/4151...

Oracle converges the AI data stack to give enterprise agents a single version of truth Matt Kimball, vice president and principal analyst at Moor Insights and Strategy, told VentureBeat the data layer is where production constraints surface first.

Data is the source of truth for agentic AI - and Oracle literally wants to live up to its name and be the 'oracle' for AI data

venturebeat.com/data/oracle-...

Meshery 1.0 debuts, offering new layer of control for cloud-native infrastructure Open-source Meshery provides a workspace to design and operate Kubernetes-based infrastructure, showing engineering teams what their infrastructure looks like, how components relate, where configurati...

Open-source Meshery provides a workspace to design and operate Kubernetes-based infrastructure, showing engineering teams what their infrastructure looks like, how components relate, where configurations conflict, and the impact of proposed changes.  www.networkworld.com/article/4150...

Ai2 releases MolmoWeb, an open-weight visual web agent with 30K human task trajectories and a full training stack Ai2's MolmoWeb is the first open-weight visual web agent to ship with its full training dataset, giving enterprise teams the ability to audit, reproduce and fine-tune a browser agent without a per-call API dependency.

Computer use /Browser use is the next big thing for AI - Basically AI that does stuff for you in your browser. The problem is that until today, no one had a full open source version, with full open source training data to see how it all works.

venturebeat.com/data/ai2-rel...

Firefox 149: Major Update Brings Built-In VPN, Split View & Productivity Tools

idolinux.com/firefox-149-has-been-rel...

Forescout brings identity-driven segmentation to multi-vendor networks The new release of Forescout's 4D platform uses a zone modeling approach to group devices, relying on device identity rather than network location.

The new release of Forescout's 4D platform uses a zone modeling approach to group devices, relying on device identity rather than network location. www.networkworld.com/article/4149... @networkworld.bsky.social

Nile adds microsegmentation and native NAC to its secure NaaS platform Updates to Nile's network-as-a-service platform include embedded microsegmentation capabilities along with native access-control features that are aimed at eliminating the need for standalone NAC appl...

Updates to Nile's network-as-a-service platform include embedded microsegmentation capabilities along with native access-control features that are aimed at eliminating the need for standalone NAC appliances. www.networkworld.com/article/4147...

Versa extends SASE platform with Inbound SSE and Secure Enterprise Browser New capabilities include a browser-native security layer that enforces VersaONE policy controls directly within browser sessions, and Inbound SSE, which inspects connections originating from the inter...

New capabilities include a browser-native security layer that enforces VersaONE policy controls directly within browser sessions, and Inbound SSE, which inspects connections originating from the internet and destined for enterprise applications. www.networkworld.com/article/4147...

Linux Foundation wants to shield FOSS devs from AI bug slop : Big Tech donates $12.5 million to get things rolling

AI slop bug reports have especially been plaguing open source projects with bug bounties: Linux Foundation wants to shield FOSS devs from AI bug slop buff.ly/SbVnVW1

Enterprise AI agents keep operating from different versions of reality — Microsoft says Fabric IQ is the fix Microsoft has made Fabric IQ's business ontology accessible via MCP to any AI agent from any vendor, targeting the core reason multi-agent systems fail in production: agents operating from conflicting definitions of the same business data.

venturebeat.com/data/enterpr...

Microsoft just made Fabric IQ's business ontology queryable by any AI agent via MCP — providing a new type of shared context layer.

oh and analysts have some questions...AND lots of insight in this one including

Just published on FOSS Force: Is Manjaro Done? Stick a Fork in It fossforce.com/2026/03/is-m...

Agents need vector search more than RAG ever did Qdrant's $50M Series B and version 1.17 release make the case that agentic AI didn't simplify vector search — it scaled the retrieval problem up. Here's what production deployments from GlassDollar and &AI reveal about when purpose-built retrieval becomes necessary.

Qdrant just raised $50M and what it means is that vector database (or what Andre Zayarni prefers to call the '..information retrieval layer for the AI age') is more relevant than ever before.

venturebeat.com/data/agents-...

FluidCloud’s Large Infrastructure Model targets the multicloud networking gap AI startup FluidCloud launches a Large Infrastructure Model to simplify multicloud networking, automate Terraform translation, and enable faster, safer cloud migrations across providers.

AI startup FluidCloud launches a Large Infrastructure Model to simplify multicloud networking, automate Terraform translation, and enable faster, safer cloud migrations across providers. www.networkworld.com/article/4144...

The team behind continuous batching says your idle GPUs should be running inference, not sitting dark Meta description (SEO/AEO) FriendliAI — founded by the researcher behind continuous batching, the technique at the core of vLLM — is launching InferenceSense, a platform that fills idle neocloud GPU capacity with paid AI inference workloads and splits the token revenue with operators. The company claims 2–3x the token throughput of a standard vLLM deployment.

FriendliAI's founders wrote the research inside vLLM. Now they want neocloud GPUs earning more money between workloads

"The cloud vendors, neoclouds, have GPUs but they can't really use GPUs all the time at full capacity," Byung-Gon Chun told VentureBeat.

venturebeat.com/infrastructu...

F5 brings new visibility and AI controls to Big-IP, NGINX F5 introduces new AI security controls, observability features, and NGINX enhancements to secure and manage modern application traffic.

F5 introduces new AI security controls, observability features, and NGINX enhancements to secure and manage modern application traffic. www.networkworld.com/article/4143...

Eridu exits stealth with $200M to rebuild AI networking AI networking startup Eridu launches with $200M in funding to redesign AI data center networking using a clean-sheet network switch and customer silicon.

AI networking startup Eridu launches with $200M in funding to redesign AI data center networking using a clean-sheet network switch and customer silicon. www.networkworld.com/article/4143... @networkworld.bsky.social

Databricks built a RAG agent it says can handle every kind of enterprise search Databricks' KARL agent uses reinforcement learning to generalize across six enterprise search behaviors — the problem that breaks most RAG pipelines.

Databricks' KARL solves the problem that breaks most enterprise RAG. The key lies with some advanced techniques for Reinforcement Learning (RL).

venturebeat.com/data/databri...

ServiceNow resolves 90% of its own IT requests autonomously. Now it wants to do the same for any enterprise ServiceNow's new Autonomous Workforce framework inherits enterprise permissions from deployment — so AI specialists can't exceed scope or self-escalate. CVS Health's CISO explains why governance architecture, not capability, is the real unlock for enterprise agentic AI.

ServiceNow is serious about automation. Apparently the key isn't about being smarter, it's about having the right roles, permissions and use cases.

"Don't chase butterflies. Focus on gritty, unsexy, operational use cases..."

venturebeat.com/orchestratio...