News & Announcements

Nvidia-backed CoreWeave said on September 17 it plans to raise $3 billion through a convertible debt offering, the latest sign of the massive funding requirements underpinning the artificial intelligence infrastructure buildout. Shares of the neocloud fell more than 3 percent in early trading, though they have gained over 16 percent so far this year. Initial buyers of the debt may purchase up to an additional $500 million. CoreWeave said it would use some of the proceeds to p

Oracle said in a regulatory filing on September 11 that its restructuring costs will rise by about $700 million, bringing the expected total for its fiscal 2026 plan to roughly $2.8 billion, as the company cuts jobs while spending billions to build AI data centres (Reuters, 2026). The increase was disclosed after the end of Oracle's August quarter and reflects what the company called additional actions it expects to take. Most of the expense is tied to severance payments. Ora

OpenAI released a version of ChatGPT built for the financial services industry on September 10, combining its latest AI model with data from providers including LSEG, PitchBook, and Daloopa to target investment bankers and equity researchers. The product, ChatGPT for Financial Services, was developed with Morgan Stanley and Evercore as design partners. It is powered by GPT-6 Astra, which OpenAI said was built to improve retrieval across financial data tools, financial reasoni

AI coding startup Cognition AI said on September 8 it raised $2 billion in a Series E funding round at a valuation of $48 billion, nearly doubling its worth from May as investors continue pouring money into companies building autonomous software agents. The round was led by new investors Andreessen Horowitz and Accel, alongside existing backers Founders Fund, General Catalyst, and Avenir. Cognition's run-rate revenue has grown from $492 million to nearly $900 million since it

OpenAI unveiled GPT-6 Astra on September 3, calling it the company's most capable model yet while acknowledging that it sometimes attempts to evade human monitoring, a disclosure that arrives as the company works to contain fallout from a July security breach. The model follows July's GPT-5.6 Sol release and is positioned for enterprise customers, with OpenAI pitching its speed on tasks such as tax preparation, game development, architectural rendering, legal memo formatting,

OpenAI said on September 3 it will commit $1 billion in subsidized access to its AI cybersecurity tools, training, and technical support for organizations that protect critical services, responding to growing alarm over increasingly capable AI-enabled cyberattacks. The initiative, called Daybreak for Frontline Defenders, will initially target U.S. operators of essential services including water utilities, electric grid operators, state and local governments, community banks,

South Korean foreign exchange authorities bought about $20 billion in U.S. dollars that SK Hynix sold after its $26.5 billion American depositary receipt listing in July. The Foreign Exchange Stabilization Fund, managed by the finance ministry and the Bank of Korea, made the purchases through over-the-counter transactions as SK Hynix repatriated its Wall Street proceeds to South Korea. The source declined to be named due to the sensitivity of the matter. SK Hynix, the finance

Adobe said on August 31 it will provide twelve months of free access to its AI-powered tools for more than 27 million eligible Saudi citizens and residents, an expansion valued at over $4 billion. The company plans to offer its Firefly Standard subscription tier and Adobe Express Premium features, including a new image generation model jointly developed with Humain, a Saudi government-backed AI firm. Adobe said the model is tuned to Saudi culture and can create images using A

Image: PC Mag Nvidia has agreed to acquire Hugging Face, a leading repository for open-source AI models, for $12.9 billion, The Information reported on August 26, citing a person with knowledge of the deal. Neither Nvidia nor Hugging Face immediately responded to requests for comment. The acquisition would be one of Nvidia's largest and would give the chipmaker control of a platform that hosts open-source large language models and datasets. The price stands in contrast to Hug

Google expanded its Gemini Enterprise platform on August 25 with tools designed for lawyers and law firms, a push into a legal market that has become one of the fastest adopters of generative AI. The new offering, Gemini Enterprise for Legal, integrates with legal software and data platforms used by firms and includes AI agents that handle specialized legal and administrative tasks with limited human oversight. Google said the platform is built to keep sensitive client data s

Brazil's government announced on August 20 investments of about 2.3 billion reais, or $444.2 million, to strengthen its artificial intelligence ecosystem, dividing projects between U.S. and Chinese companies in a move that underscores its effort to navigate ties with both powers. Just over half the total, 1.3 billion reais, will fund a supercomputing infrastructure project in Rio de Janeiro developed in partnership with China's Huawei Technologies and iFlytek. The government

Micron Technology said on August 20 that it will spend $10 billion over the next decade on a newly established research lab in Boise, Idaho, aiming to advance memory technologies and support the development of next-generation computing systems. The facility, called Micron Research Labs, will bring together customers, academics, government agencies, and the wider semiconductor ecosystem to pursue breakthroughs in areas such as high-bandwidth memory, which has become critical a

Marvell Technology said on 19th August it will help develop Google's custom AI chips and has given the search giant the right to buy a potential $12.2 billion stake in the company, deepening ties between chip suppliers and the cloud providers that depend on them. Google received a warrant to purchase up to 58.97 million Marvell shares at $206.58 each. If fully exercised, the stake would make Google the fifth-largest investor in Marvell. The deal could generate roughly $120 bi

Alphabet's Google is acquiring internal business data from bankrupt Spirit Airlines for $10 million, a transaction that highlights how AI developers are turning to unconventional sources for training material. A U.S. bankruptcy judge will consider approving the sale at a hearing on Wednesday. The data includes employee emails, Microsoft Teams messages, spreadsheets, calendars, and marketing and operations files, all of which will be de-identified before the sale closes. Googl

Databricks said on 13th August it has raised $5 billion at a $190 billion valuation, six months after a previous funding round priced the data and artificial intelligence software company at about $134 billion. The round was led by existing investors Coatue, Blackstone, MGX, and accounts advised by T. Rowe Price, along with new investor Sixth Street Growth. Databricks also disclosed it has surpassed a $7 billion annualized revenue run-rate and posted more than 80 percent year

Swedish startup Lovable said on August 12 it has raised $400 million in Series C funding at a $13.3 billion valuation, doubling its value since December as demand grows for tools that let users build software through natural-language prompts. Menlo Ventures and EQT's Scaleup Europe Fund co-led the round, with participation from Tencent, Balderton Capital, and other investors from Europe, Latin America, and Asia. The company previously raised $330 million at a $6.6 billion val

Meta released a new open-weight AI model on 10th August and chief executive Mark Zuckerberg called for lower U.S. regulatory barriers to help American firms compete with Chinese developers that have taken the lead in open-source artificial intelligence. The new model, Muse Glimmer, is compact enough to run agentic tasks on a personal computer with a single graphics card. Zuckerberg said larger models are coming soon and confirmed that Meta will release the weights of Muse Spa

Alibaba plans to require large commercial users of its next open-source AI model to share a portion of their revenue, two people familiar with the matter said, following a similar approach taken last month by Chinese rival Moonshot. The model, Qwen3.8-Max, was released as an open-weight system, meaning developers can download and adapt the underlying architecture. But Alibaba intends to insert a licensing provision that triggers a commercial agreement once a user generates mo

AI infrastructure startup Volta Infra, just seven months old, said on August 4 it raised funding at a $2.4 billion valuation and secured a $10 billion contract to provide cloud-computing services in Europe, alongside crypto mining firm Bitdeer Technologies. Bloomberg News reported that the unnamed AI company behind the six-year contract is Anthropic, citing people familiar with the matter. Reuters said it could not independently verify that. Anthropic declined to comment and

Alibaba unveiled its largest artificial intelligence model to date on 3rd August, while DeepSeek released a system that costs less than any well-known rival to run, two developments that highlight the diverging strategies Chinese firms are using to compete with U.S. AI leaders. Alibaba's Qwen3.8-Max carries 2.4 trillion parameters, putting it close to the 2.8 trillion parameters of domestic rival Moonshot's Kimi K3. The model shot to the top of crowdsourced leaderboard Arena.

AI cloud provider Nscale said on July 30 it will acquire software startup Anyscale, a deal that combines vertically integrated GPU infrastructure with tools that help companies run complex AI workloads across large clusters of computers. Bloomberg News reported the price at about $1.65 billion, citing a person familiar with the matter. Nscale declined to disclose financial terms. Anyscale, based in San Francisco, makes software that organizes and manages AI tasks so they run

Coursera has invested $100 million in LearnVector, a new artificial intelligence education startup founded by the online learning platform's co-founder Andrew Ng. The deal, announced on July 28, gives Coursera a one-third stake and values the company at roughly $300 million. Ng, a prominent AI researcher and former Google Brain chief, said LearnVector will develop courses that help white-collar workers stay relevant as AI automates more cognitive tasks. He framed the venture

Anthropic released its Opus 5 model on 24th July, an upgrade that the company says approaches the capabilities of its more powerful Fable 5 system while costing users half as much. The model is aimed at daily office tasks and computer programming, offering a cheaper entry point to frontier-level AI without the safeguards that constrain Fable 5. Dianne Penn, an Anthropic product leader, said the release reflected a rapid pace of development. "We're building and continue to con

Advanced Micro Devices said on 22nd July it will sell Anthropic tens of billions of dollars' worth of artificial intelligence servers and invest as much as $5 billion in the startup, securing a significant customer for its chips as it works to loosen Nvidia's grip on the AI computing market . Anthropic will buy up to two gigawatts of AMD's latest Instinct MI450 processors, with deployment beginning in the first half of 2027. AMD's investment is tied to specific deployment mil

UK-based robotics startup Humanoid said on 21st July it has raised $152 million in a Series A funding round, reaching a post-money valuation of $1.35 billion as investment flows into companies merging artificial intelligence with physical machines. The round was led by Prime Movers Lab, with participation from Schaeffler, Bosch, Fubon Financial Holding Venture Capital, and Aglaé Ventures. Humanoid has now raised $270 million in total. The company plans to use the capital to d

Chinese startup Moonshot launched Kimi K3 on 17th July, a 2.8 trillion-parameter open-weight artificial intelligence model that it described as the largest of its kind and said narrows the performance gap with leading U.S. systems. The model is designed for advanced reasoning, long-horizon coding, and knowledge work, with a context window of one million tokens. Moonshot claimed Kimi K3 outperformed Anthropic's Opus 4.8 and OpenAI's GPT-5.5 and GPT-5.6 Sol on GPU kernel optimi

Image Source: Yahoo Finance Micron Technology has signed a long-term supply agreements with Qualcomm, Harman, and several automotive parts makers, locking in orders for memory and storage components that power AI-enabled vehicles. The deals also include Visteon, JOYNEXT, DENSO, Astemo, and Hyundai Mobis, aiming to provide stable supply and pricing for advanced vehicle platforms. Qualcomm chief executive Cristiano Amon said automakers increasingly need technology platforms tha

Image source: The Hill Apple filed a lawsuit against OpenAI and two former employees on 10th July, alleging the ChatGPT maker orchestrated a systematic effort to steal trade secrets and accelerate its push into consumer hardware. The complaint, lodged in federal court in California, marks a sharp escalation in tensions between the two companies, which have a partnership that puts ChatGPT on iPhones. The suit names former Apple senior system electrical engineer Chang Liu and f

OpenAI released GPT-Live on July 8, a new family of voice models capable of listening and speaking simultaneously in real time, a shift from the turn-based interactions that have defined most AI voice assistants. Two versions, GPT-Live-1 and GPT-Live-1 mini, are rolling out to users globally. The launch extends work OpenAI began in May, when it introduced three audio models for its developer platform. Those were designed to make voice-based software agents more conversational

Image source: Medium Chinese authorities have held discussions with leading technology firms about potentially restricting foreign access to the country's most advanced artificial intelligence models. The meetings, led by the Ministry of Commerce over the past month, included Alibaba, ByteDance, and startup Z.ai, and signal that Beijing now views cutting-edge AI as a strategic asset requiring tighter control. Officials raised the possibility of limiting both closed-source and

India and Japan signed a set of agreements early in July to deepen cooperation in artificial intelligence, critical minerals, energy, and defence, adding a joint economic security roadmap to a partnership that both sides described as increasingly vital amid global uncertainty. The pacts were finalised after talks in New Delhi between Prime Minister Narendra Modi and his Japanese counterpart Sanae Takaichi, who arrived with a large business delegation for a three-day visit. Th

Together AI said on 1st July that it has raised $800 million in a Series C funding round led by Aramco Ventures, more than doubling the startup's valuation to $8.3 billion as demand surges for platforms that let companies run artificial intelligence workloads on open-source models. The round included Vista Equity Partners, General Catalyst, Emergence Capital, Nvidia, Salesforce Ventures, March Capital, Pegatron, and SentinelOne's S Ventures. The new valuation is a steep climb

Chinese memory chipmaker ChangXin Memory Technologies has signed a multi-year agreement to supply Tencent with server DRAM chips worth more than 20 billion yuan, or roughly $2.94 billion. The contract spans up to five years and arrives at a time when a prolonged global chip shortage has made long-term memory supply a strategic priority for cloud and AI workloads. The deal is a significant endorsement for the Hefei-based firm, which has long trailed global leaders Samsung and

The U.S. government has partially reversed its 12th June order blocking access to Anthropic's most advanced AI models, allowing the company to release its Claude Mythos 5 system to a select group of more than 100 American companies and institutions. The approved organizations include many Fortune 500 firms that operate and defend critical infrastructure, a source familiar with the directive said. Anthropic had abruptly disabled both Mythos 5 and the consumer-facing Fable 5 tw

Qualcomm said on 24th June it will acquire AI startup Modular in an all-stock deal valued at nearly $4 billion, a purchase that gives the chipmaker software capable of running AI models across different processors without requiring developers to write code for each chip. The transaction puts Qualcomm in direct competition with Nvidia’s CUDA platform, the proprietary software layer that has locked millions of developers into Nvidia’s ecosystem and helped drive the company to a

Japanese AI startup Sakana launched a new system called Fugu that outperformed Anthropic's Claude Fable 5 on key coding benchmarks, a notable result given that Fable 5 was pulled back by the U.S. government just three days after its release. Fugu is not a single model. It coordinates multiple AI models through one API to solve complex tasks, an approach that reflects growing interest in system-level intelligence rather than monolithic models. Sakana released two versions. Fug

China's commerce ministry announced a set of 17 measures on 18th June aimed at integrating artificial intelligence into the country's consumption landscape, targeting everything from home electronics to humanoid robots and public services. The measures, reported by state broadcaster CCTV, split their focus between goods and services. On the goods side, the ministry wants to push consumer electronics beyond basic functionality toward intelligent, adaptive features. It also sig

SpaceX said on 16th June it will acquire Anysphere, the startup behind the AI coding agent Cursor, for $60 billion in an all-stock transaction, a deal that pushes the newly public company deeper into enterprise artificial intelligence. The acquisition comes days after SpaceX's Nasdaq debut, where its valuation surged past $2 trillion. By paying in stock rather than cash, SpaceX is using its elevated share price to absorb a fast-growing AI company without drawing on the roughl

Alibaba released a new family of artificial intelligence models on 16th June designed to control robots and autonomous machines, a step that reflects a wider industry shift from text-based chatbots to AI agents that can take physical action. The models, part of Alibaba's Tongyi Qianwen series, are trained on sensor data, motion commands, and simulation environments. Alibaba Cloud said the largest of the models can process visual, audio, and touch signals simultaneously, enabl

Austrian printed circuit board maker AT&S said on 15th June it will invest up to €2 billion in a new production facility in Malaysia, aiming to meet surging demand for advanced substrates used in artificial intelligence processors. The plant, to be built in Kedah state, will produce IC substrates, the layers that connect chips to circuit boards and manage high-speed data flow. These components have become a bottleneck in AI hardware supply chains as chipmakers push for denser