Ai Chip Demand
The latest Ai Chip Demand coverage — news, analysis, and updates from the WindowsNews.AI desk.
Italy's Antitrust Watchdog Opens Probe into Microsoft Over Covert 365 Copilot Bundling and Price Surge
Italy's competition authority, AGCM, has opened an antitrust investigation into Microsoft over alleged unfair commercial practices linked to the automatic inclusion of Copilot AI features in Microsoft 365 subscriptions, which led to price increases without clear user consent. The probe could result in fines and force changes to how AI tools are bundled and sold across Europe, potentially setting a precedent for consumer rights in the age of AI.
Italian Antitrust Regulator Opens Probe Into Microsoft Over Forced Copilot Bundling and 365 Price Hikes
Italy's AGCM launched an investigation on June 26, 2026 into Microsoft's bundling of Copilot AI with Microsoft 365 subscriptions, alleging unfair pricing practices that limit consumer choice. The probe could result in fines and force Microsoft to offer unbundled plans, setting a precedent for AI-regulation in the EU.
NPUs Storm Windows PCs: The 2026 AI Chip Showdown Redrawing Cloud Economics
By 2026, NPUs will be standard in Windows laptops, running AI inference locally and slashing cloud costs. Meanwhile, GPUs still dominate training, TPUs anchor Google’s AI stack, and ASICs offer hyperscale efficiency. The result is a four-way chip battle that transforms Windows PC architecture and enterprise AI spending.
Daloopa Unveils MCP Connector for Microsoft 365 Copilot, Bringing Source-Linked Financial Data to Investment Pros
Daloopa has launched an MCP connector for Microsoft 365 Copilot, giving investment professionals direct, source-linked access to financial data on over 5,500 companies. The integration reduces AI hallucination risks and enables audit-ready workflows across Excel, Word, and PowerPoint, while aligning with Microsoft’s open MCP standard for Copilot extensibility.
Intel Stock Surges After Trump Reveals Apple Partnership for U.S. Chip Manufacturing
Intel's stock surged on June 18, 2026, after President Trump announced a potential Apple partnership for U.S. chip design and manufacturing. While details remain unconfirmed, the news has ignited optimism about Intel's foundry turnaround, though significant challenges persist before a deal can materialize and deliver on its promise.
Shadow AI Epidemic: 76% of Workers Already Bring Their Own AI to Work
A May 2026 survey reveals that 76% of workers use their own AI tools at work, even though 41% report receiving no AI tools, training, or guidance from their employer. This article explores the risks of shadow AI on Windows devices and outlines governance strategies for IT leaders.
Adobe's Topaz Labs Deal Brings Cutting-Edge AI Video Upscaling to Windows Creators
Adobe announced on June 25, 2026, that it has signed a definitive agreement to acquire Topaz Labs, integrating its AI-powered video upscaling, denoising, and stabilization tools directly into Creative Cloud apps like Premiere Pro and After Effects. The move promises to streamline workflows for Windows users, leveraging NPUs and GPUs for enhanced performance. While community reactions are mixed over potential subscription changes, the acquisition positions Adobe as a leader in AI-driven video post-production.
PitchBook's Federated Connector Brings Private Markets Data to Microsoft 365 Copilot and Excel
PitchBook's new federated connector integrates private market intelligence directly into Microsoft 365 Copilot, enabling users to query and analyze data within Excel, Chat, and other apps using natural language. The integration eliminates manual data transfers and enhances financial workflows for private equity, venture capital, and related fields. Available to PitchBook subscribers with Microsoft 365 Copilot, the connector marks a significant step toward AI-driven financial analysis.
AMD FSR 4.1 Tested on Radeon RX 7000: Better Pictures, Slower Frames
AMD has backported its AI-driven FSR 4.1 upscaling technology to Radeon RX 7000 series GPUs via a driver update. While the new version offers markedly improved image quality, independent testing indicates a performance penalty compared to the non-AI FSR 3.1 on RDNA 3 hardware. Gamers must now weigh the visual benefits against frame rate costs, though future driver optimizations may narrow the gap.
Italian Antitrust Watchdog Targets Microsoft Over Unexplained AI Bundling in 365 Price Hikes
Italy's competition authority AGCM has launched an investigation into Microsoft over allegations it failed to adequately explain price increases for Microsoft 365 tied to the mandatory inclusion of AI-powered Copilot and Designer tools. The probe examines whether consumers were given clear notice and genuine choice, with potential fines and remedy orders that could reshape how AI subscriptions are marketed across the EU.
June 25 Copilot for Excel update adds reusable Skills, ERP connectors, and audit tracking
Microsoft's June 25, 2026 update to Copilot in Excel introduces reusable Skills for automating repetitive tasks, native finance data connectors for ERP systems, planning controls for scenario modeling, and detailed change attribution for auditability. The features roll out across Excel for web, Windows, and Mac, aiming to boost productivity for finance professionals while addressing trust and compliance concerns.
AMD Surprises Gamers with Early FSR 4.1 Support on Radeon RX 7000: Adrenalin 26.6.2 Driver Delivers INT8 Upscaling Boost
AMD has released FSR 4.1 support for Radeon RX 7000 GPUs ahead of schedule via the Adrenalin 26.6.2 driver, introducing an INT8 inference model for AI-accelerated upscaling. A follow-up 26.6.3 hotfix is poised to address Windows 10 stability issues, underscoring AMD’s aggressive push to match DLSS 4 performance while maintaining broad platform support.
Nvidia’s 20-Year CUDA Moat: Why Raw Benchmarks Can’t Topple the AI GPU King
Nvidia's two-decade investment in the CUDA ecosystem has created an insurmountable moat in AI GPU computing, even when AMD offers faster hardware on paper. Windows users face a stark reality: CUDA's seamless toolchain, exhaustive library support, and massive developer community make Nvidia the only viable choice for machine learning, despite AMD's strides with ROCm. Raw benchmark numbers simply cannot overcome the network effects of software lock-in.