TL;DR

OpenAI targets $6B sale at $500B valuation; SoftBank buys Ohio AI plant; China mandates domestic AI chips.

Highlights

• OpenAI targets a $6B secondary share sale at a $500B valuation; SoftBank, Thrive, Dragoneer reviewing 1.
• SoftBank acquires Foxconn’s Ohio plant for $375M to manufacture AI servers, launching “Stargate” with OpenAI and Oracle 4.
• China mandates >50% domestic chip use in data centers; Cambricon raises $560M for AI chips; Nvidia resumes H20 shipments 2.
• Salesforce acquires AI automation startup Regrello (terms undisclosed) to expand process automation tools 6.
• Google increases TeraWulf stake to 14% with $1.4B backstop; TeraWulf expands data center capacity 5.
• Applied Digital plans $3B AI-focused data center in North Dakota, targeting 280MW initial capacity 5.
• Dayforce surges 26% on Thoma Bravo takeover talks, indicating ~$7B valuation 7.
• Gemini (Winklevoss-backed) confidentially files for Nasdaq IPO amid crypto market volatility 8.
• Oracle integrates OpenAI’s GPT-5 across cloud/database products; OpenAI releases open-weight GPT-5 models 13.
• Anthropic enables Claude AI to terminate abusive chats, tightens safety policy, and eases some political restrictions 14.
• Meta reorganizes AI division for the fourth time in six months; shares fall 3% 15.
• Paradigm raises $5M seed led by General Catalyst, launches AI-powered spreadsheet platform to public 10.
• Amdax plans Euronext listing for bitcoin treasury vehicle targeting 1% of bitcoin supply (~$24B) 12.
• Soho House to go private in $2.7B deal led by MCR Hotels; Apollo provides $700M in funding 3.
• Google and Kairos Power to build a 50MW advanced nuclear reactor for TVA grid, supplying Google data centers 9.
• The Argument, a digital magazine, launches with $4M seed at $20M valuation 11.

Commentary

OpenAI’s planned $6B secondary sale at a $500B valuation cements its position as the most valuable private AI company, attracting significant late-stage capital and setting a high-water mark for sector valuations 1. The deal, alongside SoftBank’s $375M acquisition of Foxconn’s Ohio plant to manufacture AI servers for the “Stargate” initiative, signals continued investor demand for both AI software and supporting infrastructure 4. These moves reinforce the trend of vertical integration among major AI players and highlight the capital intensity required to scale in this space 14.

China’s mandate for over 50% domestic chip use in data centers, Cambricon’s $560M raise, and Nvidia ’s resumed H20 shipments illustrate the ongoing bifurcation of the global AI hardware market 2. This creates both opportunity and risk for VCs: domestic champions in China are likely to see increased funding and M&A, while U.S. and European startups may face new export and competitive headwinds 2.

Strategic M&A remains active: Salesforce ’s acquisition of Regrello demonstrates continued demand for AI-native automation in enterprise software 6, while Dayforce and Soho House are both in play for private equity takeouts, reflecting ongoing interest in public-to-private transactions 73. In crypto, Gemini’s confidential IPO filing and Amdax’s Euronext plans show that institutionalization of digital assets is progressing, even as market volatility persists 812.

The infrastructure buildout continues with Google ’s increased stake in TeraWulf 5, Applied Digital ’s new AI data center 5, and Google ’s partnership with Kairos Power on advanced nuclear—each addressing the surging power and compute needs of AI workloads 9. Meanwhile, product innovation and governance remain front and center: OpenAI’s GPT-5 integration with Oracle 13, Anthropic’s new safety features 14, Meta’s ongoing AI division reorganizations 15, and the launch of Paradigm’s AI spreadsheet reflect both the pace of technical change and the importance of responsible scaling 10.

For VC investors, the environment remains highly competitive at both early and late stages. Key areas to watch: late-stage AI deal pricing, cross-border chip and infrastructure dynamics, enterprise AI M&A, and the evolving regulatory and governance landscape for advanced models.