OpenAI plans to spend about $100 billion renting backup servers from cloud providers over the next five years, according to reporting shared with shareholders and summarized by Reuters. The move underscores the scale of redundancy and capacity OpenAI targets as model usage and uptime requirements surge globally.
Meta unveiled its first consumer-ready smart glasses with a built-in display at its Connect event, extending momentum from its Ray-Ban line into an on-lens HUD experience. The launch sits alongside wristband gesture control and an ~$800 price point discussed by outlets ahead of the event, signaling a premium push for AR wearables.
Amazon is developing consumer AR glasses that would rival Meta's offerings, with features like turn-by-turn navigation shown on a small on-lens screen, per The Information reporting relayed by Reuters. The project would deepen Amazon's ambient computing strategy while raising competitive stakes in consumer AR.
Hundreds queued at Apple's Beijing flagship on iPhone 17 launch day, pointing to promising early demand in China despite a competitive market. Video footage published by Reuters showed launch-day crowds at the store as sales opened.
Apple asked at least two suppliers to raise production of the entry iPhone 17 by 30% after strong pre-orders, according to The Information via Reuters. Early demand signals could shift the product mix and channel allocations heading into Q4.
Reuters highlighted a broader wave of multibillion-dollar investments to expand AI infrastructure-spanning chips, power, and data center capacity-amid escalating model and inference demands. The Nvidia-OpenAI arrangement is a flagship example of capital concentration to secure compute and accelerate AI roadmaps.
Legal experts told Reuters the $100 billion Nvidia-OpenAI partnership raises significant antitrust issues given Nvidia's dominance in AI chips and OpenAI's market power in foundational models. Observers worry that supply-and-investment entanglements could disadvantage rivals across the stack.
A weekly AI roundup noted Gartner's forecast that worldwide AI spending will reach $1.5 trillion in 2025, reflecting enterprise urgency to operationalize AI at scale. The same digest flagged enterprise data readiness gaps from Acceldata, emphasizing integration, governance, and observability challenges.
Microsoft will pay to use Anthropic's technology for some AI features in Office 365 apps, signaling a diversification of AI partners beyond OpenAI. The shift highlights a maturing enterprise AI stack where multiple frontier model providers power distinct workloads.
A federal judge declined-for now-to approve Anthropic's proposed $1.5 billion settlement with authors alleging unauthorized training on their works. The decision sends the parties back for potential revisions as courts probe how to resolve AI-era copyright disputes at scale.
Anthropic's Economic Index finds AI usage concentrated in higher-income regions and coding/data tasks, with automation strongest where APIs enable programmatic use. The report warns productivity gains may widen inequalities without targeted policies and deployment strategies.
Deep tie-up between AI leaders raises competition concerns
Nvidia will invest up to $100 billion in OpenAI and supply data center chips, marking a deep tie-up between two of the most influential players in the AI race. Analysts say the structure could heighten competition concerns even as it accelerates compute buildouts and model progress at unprecedented scale.
The partnership comes at a time when demand for AI computing power is surging, with companies racing to secure the necessary infrastructure to train and deploy increasingly sophisticated models. This collaboration represents one of the largest investments in AI infrastructure to date.
While the deal promises to accelerate technological advancement, it has also raised eyebrows among regulators and competitors who worry about the concentration of power in the AI sector. The investment could potentially limit choices for other AI companies seeking access to cutting-edge chips and computing resources.
Apple introduced the iPhone 17 family including the ultra-slim iPhone 17 Air, alongside other hardware updates to drive upgrades ahead of the holidays. Apple's newsroom also detailed iPhone 17 Pro models with A19 Pro, major camera and battery advances, and deep Apple Intelligence integration across experiences.
Google began integrating Gemini into the Chrome browser in the U.S., with plans for agentic capabilities to execute multi-step tasks, retrieve past pages, and summarize across sites. The move follows a federal ruling that avoided a breakup while imposing limits on certain exclusivity agreements, keeping core distribution deals largely intact.
Alphabet reached a $3 trillion valuation for the first time, buoyed by renewed optimism around AI initiatives and a favorable antitrust ruling backdrop. The milestone underscores investor conviction in Alphabet's ability to monetize AI across search, cloud, and apps amid competitive and regulatory pressures.
U.S. federal agencies will be able to use Meta's Llama family as the administration expands adoption of commercial AI systems in government workflows. The approval widens procurement options and competitive dynamics for foundation models in public-sector deployments.
OpenAI appointed former xAI CFO Mike Liberatore as business finance officer to help oversee AI infrastructure and scaling economics. The hire comes amid rapid footprint expansion and complex financing structures for compute, data centers, and reliability.
xAI laid off at least 500 data annotators supporting Grok training while saying it plans to scale specialized AI tutor roles. The shift reflects evolving labeling strategies as companies balance cost, quality, and synthetic/agentic data pipelines.
Elon Musk said xAI is not raising capital after a CNBC report, reiterating the company's focus on scaling its Memphis "Colossus" supercomputer and model competitiveness. The denial comes amid investor exuberance for AI unicorns and speculation around mega-rounds.
A Reuters timeline framed Nvidia's $5B stake and co-development pact as fresh impetus for Intel's recovery under CEO Lip-Bu Tan after recent missteps. The partnership could bolster confidence in Intel's future nodes and platform roadmaps.
Databricks acquired a minority stake in Indicium, marking its first investment in a Latin American startup to deepen data and AI consulting reach in the region. The move aligns with enterprise AI rollouts that pair platforms with services to accelerate adoption.
Oracle is in talks with Meta for a multiyear AI cloud computing deal worth about $20 billion, reflecting Big Tech's scramble to lock in compute for model training and serving. The discussions underscore Meta's multi-vendor approach to securing capacity and performance.
SK Hynix completed internal certification and set up a customer production system for next-gen HBM4 as it looks to retain leadership over rivals in AI memory. The transition positions Hynix for the next phase of GPU memory scaling as model sizes and batch throughput climb.
Arm launched its next-generation Lumex cores optimized to run AI on smartphones and wearables without internet access, spanning four design tiers for efficiency and performance. The push advances edge AI capabilities as privacy, latency, and cost drive on-device inference.
The S&P 500 and Nasdaq closed at record highs as Oracle's surge on AI optimism and cooler inflation data bolstered expectations for a Fed cut. The session captured how AI infrastructure narratives are increasingly moving broader equity indices.
Industry coverage noted Verizon is building a 6G alliance with telecom heavyweights as next-gen wireless planning accelerates into standards, spectrum, and early trials. The same outlet highlighted a persistent public trust deficit in AI, framing the context for responsible deployment and regulation debates.
Workday used its Workday Rising US event to announce significant advances in agentic AI capabilities across its platform and apps. The updates illustrate how enterprise software leaders are embedding AI assistants and workflows natively to drive productivity.
Samsung's AI Forum gathered global researchers to discuss breakthroughs and strategy, with a spotlight on AI-driven semiconductor directions and vertical AI approaches near Yongin. The agenda featured leading academics and highlighted the tight coupling of foundational research and next-gen chip roadmaps.
TechRadar's archive highlights SK Hynix set to be first with HBM4 for Nvidia's Rubin AI GPU, outpacing Samsung and Micron in the next memory class race. Early HBM4 leadership would reinforce Hynix's position at the heart of AI accelerator performance scaling.
TechRadar reported Tesla scrapped its least-expensive Cybertruck variant after about five months on sale, intensifying questions about the product's positioning and demand. The move fuels debate on EV pricing, lineup complexity, and production trade-offs heading into year-end.
Researchers flagged by TechRadar found Chinese malware families flooding GitHub pages via SEO poisoning, including HiddenGh0st, Winos, and kkRAT. The campaign targets developers with malicious payloads masquerading as legitimate code resources.
TechRadar's roundup noted research progress on mitigating the problem of AI systems "forgetting how to behave safely," a core stability and alignment challenge. Such methods aim to preserve safety behaviors over time despite continued training or domain shifts.
Reuters' OpenAI topic page noted OpenAI tapped an Apple supplier to manufacture an AI device, reflecting its push into dedicated hardware. The page also referenced a non-binding deal with Microsoft that would allow OpenAI to restructure, signaling governance and partnership recalibration.
Russian resellers unveiled iPhone 17 in Moscow with a sharp rise in pre-orders despite macroeconomic headwinds, Reuters reported. The retail showings suggest strong early demand across multiple regions beyond China and the U.S.
Google Cloud outlined at least $58 billion in revenue expected to convert over the next two years from its growing backlog, with nine of the ten largest AI labs as customers. Alphabet also lifted 2025 capex plans to $85 billion to meet surging AI infrastructure demand and backlog conversion needs.
Google and PayPal announced a strategic, multi-year partnership to roll out AI-powered shopping and payment experiences across their platforms. The collaboration targets checkout, commerce discovery, and fraud reduction via AI-driven workflows.
Google headed to trial in Alexandria to avoid a forced sale of parts of its online advertising business in its latest face-off with U.S. antitrust enforcers. The case's outcome will shape the future structure of Google's ad stack and potential remedies across ad-tech markets.
Nvidia took a $5 billion stake in Intel and announced multi-generation products linking Intel CPUs with Nvidia AI/graphics chips via NVLink for higher system throughput. Analysts see the tie-up potentially boosting Intel's next-gen manufacturing credibility even without firm Nvidia foundry commitments.
Oracle said remaining performance obligations hit $455 billion and are set to surpass $500 billion soon, forecasting OCI revenue growth to $18 billion this fiscal year and to $144 billion over four years. The guidance reflects heavy AI infrastructure demand and multi-billion-dollar customer wins across training and inference.