# wants Published articles for wants. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## Extreme Networks wants the network to run AI, not just be run by it DevFeed: [Extreme Networks wants the network to run AI, not just be run by it]() Original publisher: [Read original article]() Author: Zeus Kerravala Published: 2026-10-30T19:53:09Z Content type: article Language: en Sources: [SiliconANGLE]() Topics: [AI Infrastructure](), [Local AI](), [Artificial Intelligence](), [Network Operations](), [Authorization](), [shadow AI]() Tags: [about](), [access-points](), [agent](), [ai](), [artificial-intelligence](), [compute](), [exchange](), [extreme-networks](), [first](), [generative-ai](), [heres](), [inc](), [information](), [information-technology](), [infra](), [infrastructure](), [it](), [just](), [network](), [networking](), [networks](), [news](), [run](), [the-latest](), [wants]() ### AI overview Extreme Networks is positioning the network as both an AI-managed system and local infrastructure for AI workloads. The article reports early Agent ONE Coworker results, including about 1,000 troubleshooting hours saved in its first month, while noting that only 58% of cases it opens include all evidence support needs. Planned for the first half of 2027, Exchange zones would combine a Universal Compute Platform or customer server with spare compute in installed access points and switches; applications would run locally on open-weight models by default, with cloud inference only when policy permits and with each use logged. The article describes potential uses of network telemetry beyond operations, but says the zones and demo apps remain unproven. It advises IT teams to involve network staff in AI planning, keep governance separate from agent execution, inventory discarded telemetry, assess edge capacity under changing network demand, and establish task costs before evaluating consumption-based pricing. ### Source excerpt At its recent AI Summit in Amsterdam, Extreme Networks Inc. shared early Agent ONE Coworker results and previewed Exchange zones, which put spare access points and switch compute to work as on-premises artificial intelligence infrastructure. Here's what it means and what information technology pros should do about it. Most of the conversation about AI and [...] The post Extreme Networks wants the network to run AI, not just be run by it appeared first on SiliconANGLE. ## Kernel eviction notice: What wesp.sys reveals about the Windows Endpoint Security Platform DevFeed: [Kernel eviction notice: What wesp.sys reveals about the Windows Endpoint Security Platform]() Original publisher: [Read original article]() Author: Pavel Yosifovich Published: 2026-10-26T00:00:00Z Content type: article Language: en Sources: [Elastic Security Labs]() Topics: [Endpoint Security & XDR](), [Endpoint security](), [Kernel](), [Windows](), [Post Mortem]() Tags: [about](), [binaries](), [block](), [client](), [edr](), [endpoint-protection-security](), [endpoint-security](), [kernel](), [let](), [microsoft](), [nothing](), [platform](), [preview](), [process](), [security](), [security-platform](), [show](), [user](), [wants](), [windows](), [with]() ### AI overview Microsoft's Windows Endpoint Security Platform (WESP) aims to reduce third-party EDR code in the kernel after the 2024 CrowdStrike outage. In the preview build described, Microsoft's wesp.sys is a file-system mini-filter at altitude 329500; it uses a filter communication port to send events to the user-mode espclient.dll, whose API includes event rules and filters. This design could let Microsoft maintain one kernel component while EDR services handle more work in user mode. The binaries also suggest a current limitation: espclient.dll imports FilterGetMessage and FilterSendMessage but not FilterReplyMessage, so the author infers that clients cannot yet reply with a decision such as vetoing process creation. The platform is early, undocumented, and its capabilities may change; the excerpt says user-mode blocking is a requirement for adoption, not an observed feature. ### Source excerpt Microsoft wants EDR out of the Windows kernel, and the preview binaries show wesp.sys telling a user-mode client that a process is starting, with nothing in its imports to let that client block it. ## Nvidia wants AI factories built in real factories DevFeed: [Nvidia wants AI factories built in real factories]() Original publisher: [Read original article]() Author: Frederic Lardinois Published: 2026-10-21T11:30:00Z Content type: news Language: en Sources: [The New Stack]() Topics: [NVIDIA DSX](), [NVIDIA Vera Rubin](), [Nvidia]() Tags: [2](), [2016](), [about](), [ai](), [ai-factories](), [ai-factory](), [ai-infrastructure](), [all](), [also](), [announcement](), [aws](), [berlin](), [capacity](), [conference](), [data](), [data-center](), [gtc](), [nvidia](), [show](), [up](), [wants]() ### AI overview Nvidia's DSX POD is a factory-preassembled data-center building block intended to bring new AI capacity online faster. Each shipping container holds seven Vera Rubin NVL72 racks (504 GPUs), with cabling and networking installed and tested; after delivery, crews connect the unit to its neighbors and complete burn-in tests. Nvidia says it can be ready to produce tokens in as few as 14 days, provided the site already has power, cooling and network connections--a significant caveat because those can be bottlenecks. The design uses warm coolant and dry coolers instead of water-consuming cooling towers, though operators in some climates may need chillers on the hottest days. Nvidia estimates dry cooling could save about 54 million liters of water daily in a 2-gigawatt deployment. Partners are expected to begin worldwide deliveries in early 2027. ### Source excerpt Nvidia wants new AI capacity to show up at the data center fully built. At its GTC Berlin conference on The post Nvidia wants AI factories built in real factories appeared first on The New Stack. ## UK FCA exec wants to require disclosure of 'significant AI failures and near-misses' DevFeed: [UK FCA exec wants to require disclosure of 'significant AI failures and near-misses']() Original publisher: [Read original article]() Author: Evan Schuman Published: 2026-10-21T01:05:08Z Content type: news Language: en Sources: [CIO]() Topics: [Agentic AI](), [ai security](), [LLM security](), [Finance]() Tags: [about](), [accountability](), [added](), [agent](), [agent-identity](), [agentic](), [agentic-ai](), [agentic-finance](), [agentic-systems](), [ai](), [also](), [artificial-intelligence](), [artificial-intelligence-government-laws-and-regulations-security](), [bloomberg](), [but](), [change](), [chief-data-officer](), [conditional](), [continuous](), [disclosure](), [edition-us](), [government](), [language-en](), [laws-and-regulations](), [security](), [storytype-news](), [uk](), [wants]() ### AI overview FCA Chief Data Officer Jessica Rusu called for firms to report significant AI failures and near-misses, extending existing major-incident reporting rules. She said agentic systems need continuous assurance because behavior can change at scale, under stress, or when underlying models change; the FCA plans to work with industry, regulators, and frontier labs on an agentic-finance framework covering identity, delegated authority, accountability, and execution. Malwarebytes researcher Pieter Arntz described failures as events causing serious harm and near-misses as plausible harms stopped by controls, while warning that weak penalties could make reporting a paper exercise. Other experts cautioned that stronger monitoring could make safer firms appear to have more incidents, and that reports and warnings may have little effect without clear laws and fines. They also flagged supplier model updates and outages as risks requiring scrutiny. The excerpt says the FCA had not clarified what qualifies as "significant." ### Source excerpt When UK Financial Conduct Authority (FCA) Chief Data Officer Jessica Rusu gave a speech on Monday advocating that UK companies be required to report "significant AI failures and near-misses," she added to the drumbeat of demands for some kind of regulatory control over enterprise AI deployments. But her speech also highlighted regulator frustration over futile attempts to slow down the seemingly unstoppable agentic and genAI rollouts, as reports of agent misbehavior and the lack of meaningful consequences continue to emerge with alarming frequency. Rusu, speaking to a Bloomberg UK Policy Series audience, said that firms need to invest more in cyber and operational resilience. "Our rules already require firms to report major incidents, and that must extend to significant AI failures and near-misses. As the AI Minister has recognized, near-miss reporting for agentic AI is becoming a national security priority. It is a regulatory priority too," Rusu said. "It also means thinking differently about assurance. A system that is safe at deployment may behave differently at scale, under stress, or as the models beneath it change. Agentic systems therefore require continuous, conditional assurance." She added, "the FCA will work with industry, fellow regulators, and frontier labs to develop a trusted framework for agentic finance, including clear approaches to agent identity, delegated authority, accountability and execution." One of the questions that came out of the speech is what precisely was meant by "significant AI failures and near-misses." Questions to FCA to clarify their terminology were not answered by publication time. AI failure vs near miss But Pieter Arntz, a malware intelligence researcher at security firm Malwarebytes, offered his own take on what those terms likely mean. "A significant AI failure is one with serious consequences, not simply an incorrect answer. In financial services, that could mean an AI system authorizing substantial fraudulent payments, e ## ElevenLabs wants to run your voice agent. Here's what developers give up. DevFeed: [ElevenLabs wants to run your voice agent. Here's what developers give up.]() Original publisher: [Read original article]() Author: Amanda Caswell Published: 2026-10-20T18:43:47Z Content type: news Language: en Sources: [The New Stack]() Topics: [Conversational AI](), [Agent Harness]() Tags: [agent](), [ai-agents](), [ai-engineering](), [developer-tools](), [developers](), [elevenlabs](), [first](), [heres](), [integrate](), [managed](), [max](), [mini](), [models](), [realtime](), [reasoning](), [recognition](), [run](), [speech](), [speech-recognition](), [stack](), [up](), [voice](), [wants]() ### AI overview Eleven Realtime reduces voice-agent stack work by managing orchestration, turn detection, interruption handling, and underlying models, while leaving developers unable to select or configure those models individually. Through ElevenAgents, developers can still configure prompts, voices, tools, knowledge bases, and turn-taking; Custom mode retains control over the pipeline and model choices. In Artificial Analysis tests, Max outscored Mini on spoken reasoning (98.8% vs. 75.9%) and tool-use tasks (87.1% vs. 60.3%), while Mini did slightly better on turn-taking (95.8% vs. 94.1%); latency was similar. Migration from OpenAI's Realtime API can retain its SDK over WebSockets by changing endpoint and model, but WebRTC apps must switch transport, and developers must handle 24 kHz PCM16 capture and playback themselves. Realtime API also lacks ElevenAgents platform features such as phone and WhatsApp channels, knowledge bases, and monitoring. Mini lacks Zero Retention Mode and HIPAA configurations; Max supports them in the US and EU but is unavailable in India and Singapore. The excerpt says production performance and diagnostic visibility remain uncertain. ### Source excerpt ElevenLabs launched Eleven Realtime on Tuesday, introducing two managed voice-agent models, Mini and Max, that integrate speech recognition, reasoning, turn-taking, The post ElevenLabs wants to run your voice agent. Here's what developers give up. appeared first on The New Stack. ## Introducing AI Sessions: build internal software with your whole workspace as context DevFeed: [Introducing AI Sessions: build internal software with your whole workspace as context]() Original publisher: [Read original article]() Author: Tristan Lécuyer Published: 2026-10-20T00:00:00Z Content type: article Language: en Sources: [Windmill Blog]() Topics: [internal developer platform](), [Platform Engineering](), [internal developer platform automation](), [Agent Harness](), [CI/CD](), [observability](), [GitHub]() Tags: [account](), [ai](), [ai-ai-sessions-platform-engineering-internal-tools](), [ai-sessions](), [build](), [builds](), [context](), [deployment](), [internal-tools](), [other](), [permissions](), [pipeline](), [platform](), [platform-engineering](), [platform-teams](), [production](), [sessions](), [software](), [team](), [teams](), [tools](), [wants](), [with]() ### AI overview Windmill AI Sessions let teams build internal software using workspace context and the permissions of the person opening the session, with changes reaching production through the platform team's deployment pipeline. Sessions can create scripts, flows, AI agents and full-code apps; data pipelines remain in alpha with a separate editor. Builds start as drafts and are tested after users confirm inputs. Plan mode is read-only, Ask permission (the default) waits for confirmation before write, run or deploy actions, and Yolo proceeds without asking. Teams can review changes from dev to production, or through per-session forks and additional reviews; Enterprise Edition also supports GitHub or GitLab pull requests through Git sync and can lock production to promotion-only changes. Runs and side effects are logged, and token costs are tracked by user and model. Sessions use configured provider credentials, support listed providers and OpenAI-compatible endpoints, and run on Windmill Cloud or self-hosted instances. ### Source excerpt Platform teams are responsible for production. Every other team wants to build its own tools with AI. An AI Session builds internal software with your whole workspace as context, within your permissions, and ships it through your deployment pipeline. ## Germany Wants to Lead the World in Industrial AI: How NVIDIA Will Help DevFeed: [Germany Wants to Lead the World in Industrial AI: How NVIDIA Will Help]() Original publisher: [Read original article]() Author: NVIDIA Writers Published: 2026-10-19T19:27:08Z Content type: news Language: en Sources: [NVIDIA]() Topics: [Physical AI](), [Cosmos](), [Nvidia](), [Unitree Robotics](), [High-Performance Computing]() Tags: [agile](), [ai](), [ai-adoption](), [ai-models](), [ai-native](), [ai-powered](), [ai-roadmap](), [ai-technology](), [applications](), [automotive](), [berlin](), [ceo](), [digital-twin](), [germany](), [gtc-2026](), [industrial](), [industrial-ai](), [industrial-and-manufacturing](), [industry](), [jensen-huang](), [nvidia](), [nvidia-in-europe](), [open-source](), [physical-ai](), [roadmap](), [wants]() ### AI overview German Chancellor Friedrich Merz and NVIDIA CEO Jensen Huang discussed plans to expand industrial AI use in Germany, including a government initiative with NVIDIA focused on physical AI. Huang also announced planned investments in four German startups. The article additionally reports Siemens' autonomous manufacturing work and a large quantum-computing simulation run on Europe's JUPITER supercomputer. ### Source excerpt In the cavernous hall of Kraftwerk, a former power station in Berlin, German Chancellor Friedrich Merz and NVIDIA founder and CEO Jensen Huang sat down to discuss Germany's AI roadmap, and in particular, the industrial applications of AI. Huang opened by tracing NVIDIA's longstanding relationship with German industry, beginning with gaming systems, workstations and automotive [...] ## openEuler Wants Linux to Treat an Entire AI Cluster Like One Single Machine DevFeed: [openEuler Wants Linux to Treat an Entire AI Cluster Like One Single Machine]() Original publisher: [Read original article]() Author: Edmar Pereira Published: 2026-10-13T14:02:19Z Content type: article Language: en Sources: [It's FOSS]() Topics: [Linux Kernel](), [Linux](), [Artificial Intelligence](), [Open Source Models & Datasets](), [virtual machines]() Tags: [2026](), [agentic](), [ai](), [change](), [china-focus](), [cluster](), [europe](), [like](), [linux](), [machine](), [news](), [one](), [open](), [open-source](), [open-source-summit](), [operating-system](), [os](), [scaling](), [source](), [summit](), [system](), [wants]() ### AI overview openEuler's keynote describes an OS intended to coordinate compute, memory, accelerators, and faults across a multi-node SuperPoD as one resource pool. Its SuperPoD stack builds on UnifiedBus: general apps retain POSIX access, while AI apps can use UB Service Core APIs; openEuler reports node-failure notifications within 500 ms, 20% lower RPC latency with UMDK, and 20% lower RH2D latency in KV-cache operations. For agent workloads, it presents disposable sandboxes, a four-tier KV cache with cross-instance reuse, and oGMemory for long-term memory, with performance figures reported by the project. The excerpt says this scope extends beyond mainline Linux's primarily single-host CXL memory work, but notes UnifiedBus and the optimized TaiShan and Atlas platforms are Huawei-associated; how much of the work reaches upstream Linux and vendor-neutral standards remains an open question. ### Source excerpt At Open Source Summit Europe 2026, openEuler laid out how SuperPoD OS and Agentic Scaling change what an operating system is expected to do. ## Google wants to be the gatekeeper for enterprise AI agents DevFeed: [Google wants to be the gatekeeper for enterprise AI agents]() Original publisher: [Read original article]() Author: Taryn Plumb Published: 2026-10-09T01:00:59Z Content type: news Language: en Sources: [CIO]() Topics: [Multi Agent Systems](), [ai security](), [AI Integration Strategies](), [Claude]() Tags: [agents](), [ai](), [ai-infrastructure](), [api](), [artificial-intelligence](), [artificial-intelligence-business-enterprise-google-vendors-and-providers](), [business](), [cloud](), [code](), [developer](), [edition-us](), [enterprise](), [enterprise-ai](), [environments](), [existing](), [gemini](), [go](), [google](), [google-cloud](), [governance](), [language-en](), [microsoft](), [orchestration](), [security-governance](), [storytype-news](), [vendors-and-providers](), [wants]() ### AI overview Google introduced a Gemini agent designed to coordinate work across enterprise systems, channels, and devices, including by creating task-specific subagents and selecting among models. Analysts describe the strategy as an effort to make Google a gatekeeper for enterprise AI workflows, while emphasizing that enterprises must address permissions, identity, oversight, and auditing as agents gain autonomy. ### Source excerpt Google seems to be making a strategic play to control the AI infrastructure layer, not just roll out another shiny new agent. At its Gemini at Work 2026 event this week, the tech giant unveiled what it calls a "single, universal agent" and API in Gemini that can be kicked off from a simple prompt box. Like other available agents, it is plugged into skills, tools, and enterprise systems, can answer questions, create media, and write and run code, return finished work via of documents, inboxes, and developer environments, and spin up sub-agents to help it do its job. But Google isn't just looking to go toe-to-toe with existing offerings, or trying to convince enterprises its models are qualitatively and quantitatively superior, said independent tech analyst Carmi Levy. "Instead, it's trying to establish itself as the gatekeeper of the new enterprise operating system, powered by whatever underlying model makes the most sense," he said. 'Omnipresent access,' orchestration capabilities Google says its Gemini agent works autonomously from a single interface, and can function as a personal assistant or team member, akin to a project manager or analyst. Coworker agents can perform tasks across days, sessions, and shifting responsibilities; they have dedicated identities, and are given their own @agents.company.com emails and persistent storage. However, Google says, they only have access to specifically-defined data and channels. Enterprise-dictated security, governance, and cost controls are built-in. "You give it objectives, not instructions," Google Cloud CEO Thomas Kurian said in his keynote. "Work now starts in the prompt window." The agent runs in the cloud and has "omnipresent access," meaning it can work on nearly any device, including iOS and Android phones or Windows and Mac desktops, he said. It can also be accessed through any channel, including command line interfaces (CLIs), Google Workspace, Microsoft 365, or Slack channels, and when integrated into third-par ## The Future of SaaS/PaaS: Every Bot Wants a Bite DevFeed: [The Future of SaaS/PaaS: Every Bot Wants a Bite]() Original publisher: [Read original article]() Author: Natalie Griffeth; David King Published: 2026-08-31T00:00:00Z Content type: article Language: en Sources: [Fastly Blog]() Topics: [Bot Management](), [The Three Main Cloud Service Models](), [Web Scraping](), [Security](), [API Governance]() Tags: [2](), [2026](), [ai](), [ai-crawlers](), [api](), [bot](), [bots](), [challenges](), [crawlers](), [data](), [fastly](), [future](), [ip](), [protection](), [saas](), [security](), [tech](), [wants](), [websites](), [with]() ### AI overview In Fastly's reported SaaS/PaaS traffic, bots account for 57% and humans 43%; unwanted bot traffic is 22% above the global baseline, while human traffic is 21% below it. High Tech also sees 136% more AI bot traffic than the global baseline. The excerpt attributes the pressure to scraping and data collection, API abuse and credential attacks, origin load and cloud costs, distorted analytics, and performance risks. It recommends identifying which bots access which content and whether requests reach origin, applying controls beyond blanket blocking, considering governed monetization of AI access, and using Fastly Deception to mislead malicious automation. The proposed approach is to distinguish useful automation from unwanted activity and align access decisions with business goals. ### Source excerpt Fastly data reveals bots now outnumber humans on high-tech websites, with AI crawlers driving new security, API, and IP protection challenges. ## 7 best free open source business intelligence tools right now DevFeed: [7 best free open source business intelligence tools right now]() Original publisher: [Read original article]() Author: Ian Vanagas Published: 2026-03-24T00:00:00Z Content type: comparison Language: en Sources: [PostHog]() Topics: [bi-tools](), [business-intelligence](), [Grafana](), [semantic-layer]() Tags: [000](), [7-best](), [best](), [business](), [business-intelligence](), [data](), [enterprise](), [free](), [google](), [intelligence](), [looker](), [money](), [open](), [open-source](), [right](), [send](), [source](), [tableau](), [tools](), [wants]() ### AI overview The seven tools serve different needs: PostHog combines product analytics, SQL dashboards, and synced business data in one platform, with a hosted free tier of 1 million events and 1 million synced rows monthly. Metabase emphasizes accessible visual and SQL querying, but offers less data modeling than some traditional BI suites; its hosted plan follows a 14-day trial. Superset supports SQL databases, 40+ visualizations, and a semantic layer, while its listed AI SQL assistant is limited to Preset's paid plans. Grafana focuses on infrastructure observability, with dashboards, alerts, and a broad plugin ecosystem; its free Cloud tier includes usage limits and 14-day retention. Evidence treats dashboards and data products as code using SQL and Markdown, with AI tools in its cloud version. Redash is a lightweight SQL and dashboard tool without its own hosted service, and it has no AI query generation. Lightdash builds on dbt models and metrics, with code-oriented workflows and AI analysts, but depends on dbt and its hosted plan starts at $3,000 monthly. The text frames the choice around team needs, existing data stack, and capacity to self-host. ### Source excerpt Not everyone has money to dish out for an enterprise Tableau contract, and not everyone wants to send their data to Looker (Google) either. Free and... ## Making Websites Better through Accessibility DevFeed: [Making Websites Better through Accessibility]() Original publisher: [Read original article]() Author: andrew@nystudio107.com (Andrew Welch) Published: 2017-07-23T22:42:00Z Content type: article Language: en Sources: [nystudio107]() Topics: [Web Accessibility (a11y)](), [Website](), [software-development](), [User interface design]() Tags: [accessibility](), [accessible](), [away](), [compliance](), [customers](), [development](), [insights](), [isn-t](), [just](), [legal](), [making](), [means](), [turning](), [ux](), [wants](), [website]() ### AI overview The article explains that website accessibility helps organizations reach more people, avoid excluding customers, and support inclusive products and services. It discusses legal requirements under the Americans with Disabilities Act in the United States and recommends integrating accessibility into design and development workflows. ### Source excerpt Making your website accessible isn't just about legal compliance, it also means not turning away customers. Because who wants to do that?