AI providers

Oracle has committed to paying investors in a data centre project even if the site lacks electricity to operate, according to a Financial Times report. The arrangement highlights the growing risks and unusual financing terms emerging from the race to build data centre capacity, particularly for AI workloads, as power availability becomes a major bottleneck for cloud providers.

A filing shows ByteDance, TikTok's Chinese parent company, gained access to more than 2,000 Nvidia B200 AI chips through Nscale's data center in Norway. A Singaporean subsidiary, Spring, contributed 73% of the UK neocloud provider Nscale's 2025 revenue. The arrangement highlights how US export restrictions on advanced chips to China can be circumvented through cloud capacity hosted in third countries, raising questions among observers about enforcement of AI hardware controls.

Univer is an open-source TypeScript project from dream-num that bills itself as an 'Office Harness for AI Agents'. It provides a single runtime combining spreadsheets, documents, slides, canvas, relational tables, and PDF handling. The repository is trending on GitHub, and the framing suggests developers are interested in giving AI agents tools to create and manipulate office-style documents. Beyond the project's own description, there is little discussion in the available evidence explaining what users are saying about it.

A new research paper analyzes the privacy practices of web and mobile conversational AI agents, examining how popular chatbot services collect, share and expose user data. The study, titled 'Prompt Like a Butterfly, Sting Like a Tracker,' is drawing attention among privacy researchers and developers, with readers debating its findings on tracking, data flows and what chatbot providers do with conversation content.

Modal Labs, a startup providing AI inference infrastructure, is reportedly closing in on a $750 million funding round that would value the company at $15.75 billion, according to TechCrunch. The deal would mark a major milestone for the inference provider as demand for running AI models at scale keeps climbing. Details on investors and timing have not yet been confirmed by the company.

A Trump administration official, Gebbia, says a newly launched AI-powered government website is built using Google's Gemini and xAI's Grok models. The disclosure highlights the administration's push to bring commercial artificial intelligence tools into public services, drawing attention to the choice of multiple competing AI providers rather than a single government-developed system.

Cloud provider General Compute is adding Cerebras chips alongside its existing Nvidia GPUs, aiming to speed up AI coding agents, Fortune reports. The move reflects growing demand for faster inference hardware as developers increasingly rely on autonomous agents that need rapid, repeated model responses. It also signals rising competition for Nvidia as alternative chipmakers like Cerebras find commercial footing.

Commentary is asking where profits in the artificial intelligence boom actually land once semiconductor sales are done. The argument raised is that chipmakers capture early value, but the longer-term gains may go to software developers, cloud providers and companies that turn AI into services and products. The piece feeds a wider debate about whether today's heavy AI hardware spending will translate into durable returns across the industry.

The chief executive of SharpLink has outlined a five-year path to having finances managed by artificial intelligence. The comments, reported by Yahoo Finance, suggest the company expects AI systems to take an increasingly active role in financial management. Details on how this would work or what drove the prediction were not provided.

Anthropic has reportedly agreed to pay SpaceX nearly twice what was previously estimated for access to compute capacity, according to a report cited by Seeking Alpha. The deal signals the rising cost of securing large-scale computing power for AI development, with the space company emerging as an unexpected infrastructure provider. Details of the contract terms and timeline have not been confirmed by either company.

Featherless, a serverless AI inference provider, is making the case that heavyweight infrastructure is overkill for small, routine AI workloads. The company uses the pizza-delivery analogy to argue that many applications can be served cheaply on demand rather than keeping large GPU capacity running constantly. The argument has drawn attention among developers weighing cloud costs for machine learning deployment.

Applied Materials and Kioxia have announced a collaboration to develop memory technology aimed at artificial intelligence applications. The partnership pairs the US semiconductor equipment maker with the Japanese flash memory producer as demand grows for advanced storage to support AI workloads. Further details on the scope and terms of the agreement were not provided in the initial report.

McLean-based startup Defcon AI has appointed Margo de Naray, a veteran of the space industry, as its chief executive officer. The announcement was reported by The Business Journals. The move signals the company's intent to lean on her aerospace experience as it grows, though further details about her appointment and the company's plans were not provided.

Reports say a workforce is now directing paperwork to an AI-Zone, an area or system where artificial intelligence handles documents. The claim suggests companies are shifting routine administrative work to automated systems. Details about which organisation is involved, where the AI-Zone operates, and how the process works have not been provided.

DARPA has selected Xint to apply artificial intelligence to securing military messaging applications. The defense research agency's move points to growing interest in hardening communications used by armed forces against interception and tampering. Details on the program scope and funding were not provided in initial coverage, but the selection places Xint among contractors working on next-generation secure communications for the US military.