Monday Edition

AI News Roundup — May 31, 2026

The last day of May brought a sweeping capex story: ByteDance disclosed plans to spend up to $70 billion on AI infrastructure in 2026 as the global AI compute race surpasses $800 billion combined — while Cisco surged 20% on AI infrastructure orders, KPMG deployed Claude to 276,000 employees, Canada ruled ChatGPT violated privacy law, and Microsoft Build 2026 loomed just two days away.

ByteDance Plans Up to $70 Billion in AI Infrastructure Spending in 2026

Bloomberg reported ByteDance is planning capital expenditures of up to $70 billion in 2026, focused entirely on data centers and AI infrastructure — funded largely from its approximately $50 billion in 2025 profits. This makes ByteDance a self-funding AI infrastructure giant, a level of independence that OpenAI and Anthropic cannot claim. ByteDance's AI assistant Doubao is already the most popular AI app in China. The company is separately discussing $100 billion in capex for 2027. Combined with Amazon ($200B), Microsoft ($190B), Google ($180-190B), Meta ($125-145B), and Alibaba ($50B+), the global AI compute race is now tracking $800-900 billion in 2026 infrastructure spend.

Bloomberg · Unrot Weekly


China's CXMT Gets IPO Approval — Memory Chip Giant Posts 719% Revenue Growth

China's ChangXin Memory Technologies (CXMT) received Shanghai Stock Exchange approval for a $4.2 billion STAR Market listing on May 27 — potentially China's biggest IPO since 2022. Q1 2026 numbers are staggering: revenue up 719% year-over-year, net profit up 1,688%, with 95.73% capacity utilization. AI server DRAM now represents over 30% of revenue (up from near zero). CXMT plans to develop high-bandwidth memory (HBM), the format inside Nvidia's AI accelerators, threatening Samsung, SK Hynix, and Micron's dominance over one of AI's most strategically critical chip categories.

Unrot Weekly


KPMG Deploys Claude to 276,000 Professionals in 138 Countries

KPMG announced it is deploying Claude across its entire global workforce via its Digital Gateway platform, embedding Claude Cowork and Managed Agents into all client work across tax, legal, and advisory services. Full implementation targets September 2026 on Microsoft Azure. This follows PwC (hundreds of thousands) and Deloitte (~470,000 employees), meaning three of four Big Four firms have now standardized on Claude — creating an implicit enterprise recommendation reaching the Fortune 500 and most major governments globally. Combined, roughly 1.1 million professional services staff will have Claude access by Q4 2026.

Unrot Weekly


Cisco Surges Nearly 20% as AI Infrastructure Orders Jump

Cisco shares soared nearly 20% after reporting a surge in AI infrastructure orders — networking, switching, and security hardware that underpins data center buildouts. The result is a signal that the AI capex wave is flowing downstream beyond the hyperscalers into the enterprise networking layer. Cisco CEO Chuck Robbins called AI infrastructure demand "the strongest cycle we've seen in decades."

Third Run Time


Cohere Acquires Aleph Alpha — $20 Billion Transatlantic Sovereign AI Bet

Canadian AI company Cohere acquired Germany's Aleph Alpha, creating a transatlantic sovereign AI entity valued at approximately $20 billion. The deal targets European governments and regulated enterprises that require AI systems where data stays within their legal jurisdiction — not routed through US servers. Aleph Alpha brings the German Ministry of Digital Affairs, Deutsche Bank, SAP, and Bosch as existing clients. Cohere brings LLM depth and $1.6 billion in prior fundraising. The combined entity is headquartered in Toronto with European HQ in Berlin.

Unrot Weekly


Canada's Privacy Commissioner Rules ChatGPT Violated Privacy Law

Canada's Office of the Privacy Commissioner ruled that OpenAI's ChatGPT violated Canadian privacy law — specifically the Personal Information Protection and Electronic Documents Act (PIPEDA) — by collecting and using personal data without proper consent. OpenAI is required to implement remediation measures. The ruling is one of the first major national privacy enforcement actions against a frontier AI model and sets a precedent for GDPR-aligned regulatory actions in Europe.

Unrot Weekly


Microsoft Build 2026 Begins Tuesday — AI Agents, Copilot, and Azure AI in Focus

Microsoft Build 2026 opens June 2 in San Francisco at Fort Mason Center, with Satya Nadella and Kevin Scott delivering keynotes. Expected announcements include Azure AI Foundry updates formalizing multi-model support (adding Claude alongside OpenAI), a new AI Foundry for Windows SDK, GitHub Copilot major expansion (signaled by GitHub's Kyle Daigle presenting), and Copilot Studio upgrades for enterprise agent building. Only 2,500 physical tickets were allocated — the most concentrated Build in the conference's history — with all sessions available online free.

CNET · Unrot Weekly