AI Infrastructure

Marathon Digital Holdings announced a milestone-based deal to acquire a 1,200-acre powered land site in Matagorda County, Texas, for up to $600 million to build a digital infrastructure campus targeting 2 gigawatts of po…

Canonical's MAAS and Landscape tools together provide cloud-like automation for provisioning, patching, monitoring, and lifecycle management of Ubuntu on bare metal at scale, addressing the operational gap between Day 0 …

SpaceXAI released a new AI model, while major tech companies reshape the benchmark conversation amid a decline in AI confidence. London sees increased AI investment, with Mistral opening an office and Rubrik investing $5…

Amazon returned to the US bond market seeking at least $25 billion to fund AI spending, with initial orders peaking near $62 billion before settling near 1.6 times the deal size. The multi-tranche offering comes as Amazo…

Micron stock surged up to 9% after the memory-chip maker announced an additional $50 billion in US chipmaking spending, bringing its total planned investment through 2035 to $250 billion. The move, praised by the Trump a…

Mercor, an AI data startup founded by three college dropouts, reached $2 billion in annualized revenue in June 2025, doubling from $1 billion four months earlier. The company pays doctors, lawyers, and engineers to train…

Ollama raised $65 million in Series B funding led by Theory Ventures, bringing its total funding to $88 million, as the open-model runner now serves nearly 9 million monthly developers. The company plans to use the funds…

Microsoft Entra ID introduces Conditional Access policies for AI agent identities, enabling organizations to enforce governance based on agent classification and custom security attributes. The approach distinguishes bet…

Microsoft Entra ID custom security attributes can serve as a governance metadata layer for AI agents, enabling attribute-driven policies instead of manual per-agent management. By defining attributes like ApprovalStatus,…

OpenAI and Broadcom unveiled Jalapeño, OpenAI's first custom inference chip, on June 24, targeting 50% lower cost per token than Nvidia GPUs. The ASIC, built in nine months at TSMC's 3nm node, will not reach production u…

Token pricing for AI models faces a supply crunch and instability, with analysts debating whether foundation models will achieve sustainable pricing power or become low-margin commodity infrastructure. The market is driv…

Lithuanian data infrastructure firm Oxylabs raised $130 million from Warburg Pincus at a $3.6 billion valuation, its first outside investment since 2015. The company, which provides web data gathering tools for AI agents…

OpenAI CEO Sam Altman announced GPT-5.6 Sol, a new AI model that delivers a 54% improvement in token efficiency for agentic coding tasks, significantly reducing inference costs. The breakthrough threatens the economic as…

Microsoft Entra Agent ID and Agent 365 provide a structured governance model for AI agents in enterprise environments. The model emphasizes visibility and accountability before enforcement, with a phased approach coverin…

Tencent released Hy3, a 295-billion-parameter open-weights AI model, at roughly six cents per million input tokens, outperforming GLM-5.1 in blind expert reviews. The model features 21 billion active parameters, 256K con…

Cerebras Systems announced plans to bring its first European data center capacity online by the end of 2026 and expand to 200 MW of AI compute capacity by the end of 2027 across France and the Nordics. Part of the capaci…

Canonical launched Managed Kubeflow on Microsoft Azure, a fully managed service that runs within the customer's own cloud tenancy to eliminate the operational overhead of maintaining Kubeflow's distributed microservices,…

AI engines and autonomous agents now decide which brands to recommend, trust, and transact with, with Adobe data showing a 4,700% year-over-year increase in AI-referred traffic to U.S. retail websites and Salesforce repo…

Krumware released an experimental Epinio MCP server for Kubernetes developers on July 9, 2026, enabling AI agents to interact with Epinio API operations through tool calls. The server translates MCP tool calls into Epini…

Sven Oehme argues that the next major cost reduction in AI will come from software and system-level optimizations, not faster chips, as AI moves into production and operational economics become critical.