rsssoftware

2026年国内エージェンティックAIのコスト管理を巡るITサプライヤーの機会：フィジカルAIを見据えて

Shinya Kato — Wed, 11 Nov 2026 05:00:00 GMT

本調査レポートは、AI活用で先行する国内企業の動きから、エージェンティックAIのコスト管理を巡る需要の変化とITサプライヤーの機会を分析し、その機会のフィジカルAIへの広がりを考察している。

「エージェントの支出は、単価が下がっても総額が増え、使い方次第で大きく変わる費用になりつつある。業務1件ごとの費用と成果を突き合わせる役割は、顧客への提供価値であると同時に、成果に連動した契約を結ぶ提供側の原価管理の基盤にもなる。この記録は、AIが設備を動かすフィジカルAIでは安全に動かすための前提にもなり、ITサプライヤーにとっての機会はさらに広がる」と、IDC Japan、AI and Automationのシニアリサーチマネージャーである加藤慎也は述べている。

A Mercurial Friend: How Agentic Model Routing Can Help or Hurt Frontier AI Labs

Daeil Chun, Alessandro Perilli — Wed, 11 Nov 2026 05:00:00 GMT

This IDC Market Perspective examines agentic AI as a double-edged sword for frontier AI labs. Agentic AI drives inference demand, but its model routing and orchestration also open the door to open-weight models in hybrid architectures. Using recent IDC survey data, the document explores the forces pulling enterprise model strategy in both directions and offers advice for frontier AI labs on how to respond.

"Agentic AI is a friend to frontier AI labs, but a mercurial one. It can multiply inference demand for foundation models but can also make it easy to swap in open-weight models in hybrid architectures. But then the same hybrid architecture can also increase the use of commercial models by routing high reasoning tasks to more powerful models. Labs that compete across the hybrid architecture, rather than against it, will win the work that matters most as agentic AI matures." — Daeil Chun, senior research manager, AI Research, IDC

Palantir Posts 3Q26 Results: U.S. Commercial Revenue Overtakes U.S. Government Revenue

Alan Webber — Tue, 10 Nov 2026 05:00:00 GMT

Palantir Technologies Inc. (Nasdaq: PLTR) reported third-quarter 2026 results on November 2, 2026, with revenue of $2.313 billion, growing 96% year over year and 20% sequentially, the first quarter in which the company's revenue topped $2 billion. The U.S. business, now 84% of total revenue, grew 119% year over year, and U.S. commercial revenue surpassed U.S. government revenue for the first time. The company raised its full-year 2026 revenue guidance to $8.530 billion–$8.534 billion, representing 91% growth.

Vanta's Pillar Security Buy Tells CISOs: Stop Writing AI Policy, Start Enforcing It

Philip D. Harris, CISSP, CCSK — Tue, 10 Nov 2026 05:00:00 GMT

Vanta has signed a definitive agreement to acquire Pillar Security, adding runtime protection and adversarial testing to its Trust Platform so AI governance policies become enforceable controls rather than documents. Pillar Security, founded in 2023, gives companies one platform to discover, test, secure, and govern AI agents across their life cycle; Vanta is building native AI inventory and runtime monitoring guardrails internally for a January 2027 release, and will integrate Pillar Security's red teaming and runtime enforcement capabilities in spring and summer 2027, respectively, along with select Pillar Security inventory capabilities at a later stage. The deal is Vanta's third acquisition and lands as the company surpasses 20,000 customers and $300 million in ARR, with daily Vanta Agent usage up 253% over three quarters. For CISOs, it signals that AI governance is consolidating around platforms that can act on agent behavior in real time, not just document intended policy after the fact.

AMD's 3Q26 Results: Datacenter Revenue Reaches a Record as Venice and Helios Begin to Ship

Jeff Janukowicz — Mon, 09 Nov 2026 05:00:00 GMT

On November 3, 2026, AMD reported 3Q26 financial results. Revenue reached a record $13.377 billion, up 45% year over year and 16% sequentially, exceeding the high end of guidance. Datacenter revenue was $8.546 billion, up 97% year over year and 27% sequentially, accounting for 64% of total revenue compared with 47% a year ago. Non-GAAP gross margin was 56%, up 190 basis points year over year. GAAP EPS was $1.46, reduced by a $1.097 billion noncash charge related to the change in fair value of customer warrant liabilities.

Confluent Current 2026: IBM Sharpens the Business Case for Data Streaming as Streamhouse Targets Agentic AI

Stewart Bond — Mon, 09 Nov 2026 05:00:00 GMT

Confluent Current 2026, the first Current since IBM acquired the company, showed Confluent and IBM shifting the data streaming conversation from how the technology works to what it does for the business, anchored on a Streamhouse architecture that keeps current business context available to AI agents and applications.

This shift matters because real-time data has become a gating condition for agentic AI, and IBM's ability to explain streaming in business terms addresses a past weakness of Confluent, but the event left open where Confluent and open source streaming head next.

IBM TechXchange 2026: Governing Agents at Scale

Jim Mercer, Matthew Flug — Mon, 09 Nov 2026 05:00:00 GMT

IBM used TechXchange 2026 in Atlanta, the first held alongside HashiConf, to argue that the primary enterprise challenge in the agentic era has shifted from building agents to scaling and governing them. Announcements spanned agent identity, a cross-platform watsonx Orchestrate control plane, Terraform and Vault for agentic infrastructure, the Concert operations platform, Bob 2.0, and Z Atlas. IBM says only 25% of today's agents are enterprise ready, and its answer to scalability is a layered stack of identity, context, orchestration, and operations.

IDC MarketScape: Worldwide AI Agents for IT Service Management Software 2026–2027 Vendor Assessment

Snow Tempest — Mon, 09 Nov 2026 05:00:00 GMT

This IDC study presents a vendor assessment of the worldwide market for AI agents for IT service management software through the IDC MarketScape model.

"AI agents for IT service management are an emerging market, for which customer demands include not only agent capabilities but security, governance and accurate organizational and technical context," said Snow Tempest, research manager, IT Service Management at IDC.

IDC MarketScape: Worldwide Digital Service Providers for Large Capital Projects in Energy and Water 2026 Vendor Assessment

Gaia Gallotti, Jean-François Segalotto, Gaurav Verma — Mon, 09 Nov 2026 05:00:00 GMT

This IDC study evaluates digital service providers supporting large capital projects in energy and water worldwide. It analyzes vendors’ capabilities, strategies, and positioning in a market undergoing structural transformation as owners face a capital investment cycle that outstrips the capacity of traditional, project-by-project delivery models and move toward governing capital programs as connected portfolios. The report provides energy and water companies with a structured framework to assess service providers across capital portfolio planning, project controls, engineering data and digital thread, construction and commissioning digitalization, IT/OT integration, advanced analytics, AI, and digital twins, and platform implementation and managed services.

“The volume of capital that energy and water owners must now deploy is so vast that the operating model behind large capital projects for the past several decades is simply no longer viable, and no amount of incremental effort will make delivering one program at a time scale to the size of this investment cycle. Owners need service providers capable of restructuring how capital is governed, funded, and executed across the entire portfolio, with shared data, standardized delivery patterns, and AI proven in production, because the old way of delivering large capital projects will not survive the demands now being placed on it.” — Research Director Gaia Gallotti, IDC Energy Insights

Sprinklr's Platform Push at CXUnifiers 2026 Rests on Social Data and Customer Context

Roger Beharry Lall, Tapan Patel, Hayley Sutherland — Mon, 09 Nov 2026 05:00:00 GMT

At CXUnifiers 2026, Sprinklr launched a beta of six Marketing AI Agents and outlined a context graph and an open agent layer designed to unify its social, marketing, service, and customer feedback products into a single AI platform. The plan builds on the social and conversation data Sprinklr already collects, and its success will depend on published evidence that this context improves AI agent results and on winning over the IT and AI leaders who now take active role in choosing AI agent platforms for customer experience.