AI News for May 29, 2026 — 7min.ai

[NEWS] Anthropic closes $65B Series H at $965B valuation, eclipsing OpenAI: Anthropic raised $65B in Series H funding led by Altimeter, Dragoneer, Greenoaks, and Sequoia, valuing the Claude maker at $965B post-money and topping OpenAI's $852B mark for the first time.
[TOOLS] Claude Opus 4.8 lands with 3x cheaper fast mode and dynamic workflows: Anthropic released Claude Opus 4.8 at the same $5/$25 per-million-token pricing as Opus 4.7, but slashed fast-mode pricing 3x to $10/$50 with output running at 2.5x normal speed.
[NEWS] Apple's iOS 27 Siri overhaul gets first visual leak ahead of WWDC: Bloomberg published renders of Apple's redesigned Siri shipping with iOS 27, showing a pill-shaped chat bubble that drops from the Dynamic Island with Ask/Siri/ChatGPT options and a standalone ChatGPT-style Siri app.
[NEWS] Mistral rebrands Le Chat as Vibe, expands into industrial AI with Airbus and BMW: Mistral used its first AI NOW Summit to rebrand consumer chatbot Le Chat as Vibe, unveil a Paris-area inference data center, and announce Airbus and BMW as launch customers for a new industrial AI push covering physics simulations and manufacturing.
[NEWS] SpaceX IPO targets $1.8T valuation with 78% of cash earmarked for AI: SpaceX is targeting at least $1.8T in its upcoming IPO and plans to raise around $80B, the largest sum ever in a public listing.
[NEWS] Apollo shops $36B debt deal to buy Google TPUs for Anthropic: Apollo Global Management and Blackstone are syndicating a roughly $36B debt financing to buy Google's custom TPU chips, which Anthropic will then lease for its AI buildout.
[NEWS] Samsung ships HBM4E samples, shares jump 6%: Samsung began shipping samples of HBM4E, the industry's most advanced high-bandwidth memory, to AI accelerator customers including Nvidia.
[NEWS] Illinois passes nation's strongest AI safety law days after Trump kills federal version: The Illinois legislature passed SB 315, requiring large AI firms to publish safety plans, file annual third-party safety test results, and report critical incidents within 72 hours (24 if there's imminent risk of death).
[RESEARCH] Biohub ships ESMFold2, a 'world model of protein biology': Chan Zuckerberg Initiative's Biohub released Evolutionary Scale Models 2, including ESMFold2, a model trained on 2.8B protein sequences that claims state-of-the-art results on structure prediction and protein design.
[NEWS] EU presses Anthropic for access to frontier Mythos AI model: European Commission officials are planning meetings with Anthropic to seek information on its Mythos model and formally request making the frontier AI system available to EU member states, Bloomberg reported.

About 7min.ai · Subscribe to newsletter