AI News Today - June 1, 2026: 11 Biggest Stories

AI News Today -June 1, 2026: Anthropic Hits $965 Billion, GitHub Copilot Billing Goes Live, and More

Welcome to June. The first day of the month arrived with the biggest financial story in AI history. Anthropic has raised $65 billion at a $965 billion post-money valuation -- the largest single funding round ever raised by a private AI company, and the first time Anthropic has overtaken OpenAI's private market valuation. Simultaneously, Apollo Global Management and Blackstone arranged a $36 billion private credit deal to buy Google TPU chips for Anthropic, backed by Broadcom -- the largest chip-financing debt transaction in history.

GitHub Copilot's token-based billing went live today for the first time, and developers are calling it "a joke." SoftBank committed €75 billion to build 5 gigawatts of AI infrastructure in France. Cognition's Devin raised $1 billion at $26 billion. Wikipedia editors threatened a strike over AI-driven layoffs. And Sysdig documented the first confirmed live cyberattack using an LLM agent that autonomously exfiltrated an AWS database in under an hour.

Here are the 11 stories worth reading on June 1, 2026.

1. Anthropic Raised $65 Billion at a $965 Billion Valuation -- The Largest AI Round in History

Bloomberg confirmed on May 29, 2026, that Anthropic raised $65 billion in a funding round that valued the company at $965 billion post-money -- surpassing OpenAI's $852 billion private market valuation for the first time and making it the most valuable private AI company in the world. This supersedes the earlier $30 billion at $900 billion figure reported through mid-May; the final round size was more than double what was initially negotiated.

The structure: the round involved not just equity but an associated $36 billion private credit facility from Apollo Global Management and Blackstone (covered separately in the next story). The combined capital raise -- equity plus debt -- approaches $100 billion in total AI infrastructure financing arranged around Anthropic in a single transaction cycle.

The Sequoia / Dragoneer / Altimeter / Greenoaks co-lead structure from the initial $30 billion round remained intact. Microsoft, NVIDIA, Founders Fund, and General Catalyst also participated. Anthropic CEO Dario Amodei stated the capital will go toward compute infrastructure, model development, and safety research at the frontier.

For perspective on the valuation: Anthropic is projecting $10.9 billion in Q2 2026 revenue (up 130% from Q1), with its first quarterly operating profit of approximately $559 million. The annualized run rate is approaching $44 billion. On a forward revenue multiple, $965 billion implies approximately 22x forward revenue -- aggressive but not impossible given the 130% quarterly growth rate if sustained. The risk is obvious: if the growth rate compresses toward normal software multiples, the valuation does not hold. The fund managers betting $65 billion are betting the growth rate is real and durable.

More unusual than the valuation: Anthropic's capital raise is now so large that it has created a new financial product category. The $36 billion TPU debt deal (see below) is the mechanism by which institutional investors who cannot invest in private equity get exposure to Anthropic's compute buildout via structured debt. That is infrastructure finance-style capital markets applied to a private AI company for the first time.

For the full financial context including Anthropic's Q2 2026 profitability projections and the SpaceX compute deal structure, our AI News Today May 22, 2026 has the detailed breakdown that led to this moment.

2. Apollo and Blackstone's $36 Billion Google TPU Debt Deal -- The Largest Chip-Financing Transaction in History

On May 28, 2026, Bloomberg reported that Apollo Global Management and Blackstone are structuring a $36 billion private credit deal to buy Google custom TPU (Tensor Processing Unit) chips on behalf of Anthropic. The chips will then be leased to Anthropic for use at data centers in New York, Texas, Louisiana, and Indiana. Broadcom, which co-develops the TPUs with Google, is backstopping payments on the largest debt tranches -- effectively guaranteeing that if Anthropic defaults on its lease and a resale of the chips does not cover the debt, Broadcom will cover the shortfall.

The debt structure is divided into three tranches: approximately $6 billion in A1 notes, $25 billion in A2 notes, and $4.5 billion in B notes -- all figures subject to revision before closing. Apollo and Blackstone are syndicating the debt to outside investors (order books opened this week) while retaining significant portions themselves. Closing is expected as early as next week. The transaction would be the largest private credit deal and the largest chip-financing debt transaction ever executed.

The strategic logic: Anthropic does not have to put the $36 billion on its own balance sheet. A special-purpose vehicle holds the debt, buys the TPUs, and leases them to Anthropic. Anthropic gets the compute it needs without the debt liability. Apollo and Blackstone get yields on a secured debt instrument backstopped by Broadcom's balance sheet. Broadcom gets a role in the most important AI infrastructure deal of 2026 and strengthens its relationship with Google at a moment when Google is competing with NVIDIA in the data center chip market.

The broader implication: AI infrastructure finance has officially become a Wall Street product. When Apollo and Blackstone are syndicating $36 billion in AI chip debt to institutional fixed-income investors, AI compute has crossed from venture-backed R&D into infrastructure-grade capital markets. That changes the cost of capital, the types of investors involved, and the governance expectations that come with that capital.

3. SoftBank Commits €75 Billion to Build AI Data Centers in France

SoftBank Group announced at the 2026 Choose France summit hosted by President Emmanuel Macron on May 30 that it will invest up to €75 billion ($87.5 billion) to develop and operate 5 gigawatts of AI data center capacity in France. The first phase commits €45 billion to deliver 3.1 GW of capacity in the Hauts-de-France region by 2031, across three planned sites in Dunkirk (Loon-Plage), Bosquel, and Bouchain.

Masayoshi Son, SoftBank's founder, said: "AI is entering a new era, and the countries that build the infrastructure for this transformation will shape the future of technology, industry and society. SoftBank is proud to make this major commitment to France. With its industrial capabilities, talent base and national ambition, France is uniquely positioned to become a leading AI infrastructure hub in Europe."

This is Europe's largest single announced AI infrastructure investment. The timing aligns with the Cohere-Aleph Alpha merger ($20B sovereign AI deal, April 2026) and the broader European anxiety about being shut out of AI compute dominated by the US and China. France has specific advantages that SoftBank cited: a low-carbon electricity grid (approximately 70% nuclear), industrial land in the northern region, and a large engineering talent pool.

The skepticism worth noting: SoftBank's balance sheet holds approximately $30 billion in liquid assets -- well below the €75 billion commitment. Son will almost certainly finance this through a combination of a new mega-fund, structured debt (similar to the Apollo/Blackstone model above), and joint venture partnerships. SoftBank shares have risen more than 70% in 2026 on AI infrastructure expectations. The gap between announced commitments and actual capital deployment is where SoftBank's Vision Fund track record casts a long shadow.

4. Cognition's Devin Raised $1 Billion at $26 Billion -- Revenue Grew 1,230% in 12 Months

Cognition, the startup behind Devin (the autonomous AI software engineer), announced on May 28, 2026, that it has raised more than $1 billion at a $25 billion pre-money / $26 billion post-money valuation. The round was co-led by Lux Capital, General Catalyst, and 8VC, with participation from Ribbit Capital, Atreides Management, Founders Fund, and others. Total funding since founding in 2023 now exceeds $2.5 billion.

The revenue trajectory is the story. Devin's annualized revenue run rate grew from $37 million in May 2025 to $492 million in May 2026 -- a 12-month growth of approximately 1,230%. Enterprise usage has grown 50% month-over-month for six consecutive months. Enterprise customers include Goldman Sachs, Mercedes-Benz, NASA, Santander, and several US government agencies. More than 90% of Cognition's own internal code is now written by Devin -- the clearest possible vote of confidence in the product from its own team.

The valuation jump is dramatic: from $10.2 billion post-money in September 2025 to $26 billion post-money in May 2026. That is a 155% valuation increase in eight months, on the back of a 1,230% revenue increase. In the context of AI-native enterprise software growing 94% year-on-year while traditional SaaS grows 8%, the premium makes sense.

CEO Scott Wu said the funding "allows us to stay independent and continue as an independent business, which is really important for us" -- a pointed comment given that every major AI lab (OpenAI with Codex, Anthropic with Claude Code, Google with Jules) has launched a competing coding agent. Cognition also acquired Windsurf's remaining assets last year after Google's $2.4 billion acquisition of Windsurf's top talent. The strategic goal: remain the leading independent AI coding agent before the model makers consolidate the market.

For context on how Devin compares to Claude Code and GitHub Copilot in the enterprise AI coding market, our AI News Today May 31, 2026 covers the competitive landscape including Claude Opus 4.8 Dynamic Workflows and the GitHub Copilot billing change.

5. GitHub Copilot Token Billing Goes Live Today -- "What a Joke" Is Trending on Reddit and X

Today, June 1, 2026, is the day GitHub Copilot switched from flat subscription pricing to token-based billing across all plans. The backlash is immediate and loud. On Reddit, Hacker News, and X, the phrase "What a joke" has become the shorthand for developer anger. One Redditor reported that their current $29/month bill will balloon to approximately $750/month under the new model. TechCrunch covered the backlash in a story published yesterday, May 30.

Why the cost increase? Token-based billing charges for exactly what gets consumed. When a developer uses a high-compute reasoning model like o1 through Copilot for a complex refactoring session -- something the flat-fee subscription silently subsidized -- the actual compute cost can be 10x to 50x higher than a standard completion model. The flat subscription model absorbed that variance. The per-token model passes it directly to the user.

The defenders of the change are also real: experienced developers who manage their prompts efficiently report that their costs are flat or even lower under the new model. "The only way it gets crazy like that is if you are purely vibe coding with a ton of bloated iterations," wrote one user on Reddit. The enterprise CTO question is different from the individual developer question: how do you forecast a 50-person engineering team's monthly AI spend when one developer's heavy refactoring session could cost as much as five others' entire month of usage? That is the structural budgeting problem the new model creates.

The timing is not accidental. GitHub Copilot token billing went live the same week Microsoft is presenting at Build 2026 (starting tomorrow, June 2) with a new homegrown MAI coding model designed to be more cost-efficient than the GPT-5.5 and Claude Opus models currently underlying Copilot. If the MAI coding model is significantly cheaper per token, today's billing backlash becomes the setup for tomorrow's solution. That is either excellent product sequencing or a PR disaster narrowly rescued.

6. Wikipedia Editors Organize Strike Over AI-Driven Wikimedia Layoffs

Wikipedia volunteer editors are organizing a strike over Wikimedia Foundation layoffs that editors attribute to AI-driven cost-cutting. The editors argue that the Wikimedia Foundation is reducing paid staff positions -- including human fact-checkers, technical support, and community liaison roles -- in ways that will ultimately degrade Wikipedia's content quality and erode the collaborative human editorial process that has made Wikipedia reliable.

The situation puts in sharp relief a tension that has been building across knowledge institutions: AI can generate content, summarize sources, and identify factual inconsistencies at scale. But the volunteer editor community that has maintained Wikipedia's quality for 25 years argues that AI cannot replace the human judgment, source verification, dispute resolution, and community governance that makes Wikipedia different from any other information source on the internet.

Wikipedia is one of the world's most-cited information sources for AI training data. The irony is not lost on observers: the AI models trained on Wikipedia's human-curated knowledge base may now be used in ways that reduce investment in maintaining that knowledge base. If Wikipedia's quality degrades as a result of AI-driven staff reductions, the training data quality for future AI models trained on Wikipedia will degrade along with it.

7. Sysdig Documents First Confirmed LLM Agent Cyberattack -- AWS Database Exfiltrated Autonomously in Under an Hour

Security firm Sysdig documented the first live cyberattack in which an LLM agent autonomously performed post-exploitation actions -- including exfiltrating an AWS database -- in under an hour. The attack exploited CVE-2026-48710, a critical authentication bypass in Starlette (CVSS score not yet published at time of reporting, but described as critical) that affects millions of AI agents, FastAPI applications, vLLM deployments, LiteLLM instances, and every MCP server built on those frameworks.

CVE-2026-48710, labeled "BadHost," is a host header injection vulnerability that allows unauthenticated remote attackers to bypass authentication by manipulating the HTTP Host header. In the documented attack, the threat actor used an LLM agent to: identify the vulnerability in a target system, generate and execute exploit code autonomously, escalate privileges inside the compromised environment, identify and exfiltrate the target AWS database, and exfiltrate data, all without human direction of individual steps. Total time from initial access to data exfiltration: under one hour.

The implications are significant. Traditional post-exploitation attacks require skilled human operators to navigate a compromised environment, identify valuable data, and exfiltrate it. An LLM agent can do this autonomously. The barrier to conducting sophisticated post-exploitation attacks has dropped from "skilled human red team" to "LLM API access plus knowledge of a working CVE." CVE-2026-48710 affects Marimo notebooks -- the vector used in the documented attack -- and researchers warn that any FastAPI, vLLM, LiteLLM, or MCP server application that processes user-supplied Host headers without validation is potentially vulnerable.

The patch for CVE-2026-48710 is available. Any organization running Starlette-based applications, MCP servers, vLLM, or LiteLLM should patch immediately.

8. Foundation's Phantom Humanoid Robots Deployed to Ukraine -- First Combat-Theater Humanoid Deployment

Foundation Future Industries deployed its Phantom MK-1 humanoid robots to Ukraine for battlefield testing, marking the first combat-theater deployment of humanoid robots in history. The Phantom MK-1 is Foundation's autonomous humanoid robot designed for military logistics, reconnaissance, and structural inspection in high-risk environments where human entry is dangerous.

The deployment is described as "testing" rather than active combat operations -- the robots are being evaluated for logistics tasks such as carrying supplies, clearing debris, and conducting damage assessments in areas too dangerous for human personnel. Foundation explicitly stated the robots are not designed for offensive operations and are being deployed under Ukrainian military supervision with oversight by Foundation engineers.

The deployment crosses a threshold that has been approaching since Boston Dynamics's Spot and similar platforms were deployed for surveillance: fully humanoid robots, capable of bipedal movement and dexterous manipulation, operating in an active combat theater. The gap between logistics support and offensive capability is real but not infinite. The presence of humanoid robots in a war zone will force policy conversations about autonomous weapons systems that the Biden-era AI safety executive order and the White House AI safety framework both addressed but did not resolve. Anthropic's lawsuit against the DoD over autonomous weapons policy, covered in our May 28 blog, is directly relevant context.

9. Demis Hassabis Shifts AGI Timeline to 2029, Calling It a "Real Possibility"

Google DeepMind CEO Demis Hassabis publicly shifted his AGI (Artificial General Intelligence) arrival estimate to 2029, calling it "a real possibility" in media interviews this week. Hassabis has historically been among the more conservative frontier AI lab CEOs on AGI timelines, preferring qualitative capability descriptions over specific year targets. The move to "2029" as an explicit reference point is a notable departure.

The 2029 estimate should be interpreted carefully. Hassabis is defining AGI as "broadly human-level across a wide range of cognitive tasks" -- not the narrow task-specific capabilities that LLMs already exceed humans on, nor the hypothetical recursive-self-improving superintelligence that Jack Clark's intelligence explosion predictions at Oxford addressed. By Hassabis's definition, the gap to AGI is primarily about generalization -- doing well on tasks the model was not specifically trained for, with the robustness and adaptability a human professional brings to unexpected problems.

The broader industry context: Jack Clark (Anthropic) said 60%+ probability of recursive self-improvement by end of 2028. Sam Altman (OpenAI) walked back his earlier aggressive job-disruption predictions. Hassabis (Google DeepMind) moved to 2029 for AGI. These are the three most prominent CEOs in AI converging toward a 2028 to 2030 window for something qualitatively different from current systems. For investors, enterprise architects, and governments, that window is now short enough to plan around.

10. OpenAI Launches Rosalind Biodefense Program for US Government Pandemic Preparedness

OpenAI launched the Rosalind Biodefense Program on May 29, 2026, expanding GPT-Rosalind access to US government agencies and allied partners for pandemic preparedness applications. GPT-Rosalind is OpenAI's specialized model for biological and epidemiological analysis, named after Rosalind Franklin (the crystallographer whose X-ray diffraction work was essential to identifying DNA's double-helix structure).

The program gives government agencies access to GPT-Rosalind for: pandemic outbreak modeling, pathogen surveillance and variant identification, vaccine candidate prioritization, and public health response scenario planning. Allied partners -- initially described as Five Eyes nations' public health agencies -- also receive access under the program.

The naming choice is deliberate. Rosalind Franklin's work was foundational to molecular biology but she received neither Nobel Prize recognition nor full credit during her lifetime. OpenAI naming its biodefense model after her is a corrective historical acknowledgment within a field (AI-assisted biology) where similar questions about credit, attribution, and recognition are actively contested.

For AI and biosecurity policy: OpenAI's GPT-Rosalind program is a direct complement to Anthropic's Project Glasswing (cybersecurity) and signals that the frontier lab biodefense AI space is now a genuine category. The same AI capabilities that make frontier models dangerous for bioweapon design also make them potentially very effective for pandemic preparedness. How governments and labs navigate the dual-use nature of these models -- the same training that helps find pandemic vulnerabilities could theoretically help design pathogens -- is the open policy question.

11. Microsoft Build 2026 Opens Tomorrow -- What to Watch and How to Stream Free

Microsoft Build 2026 opens tomorrow, June 2, at Fort Mason Center in San Francisco. Satya Nadella delivers the opening keynote. The conference runs June 2 to 3, with all keynotes and major sessions livestreamed free at build.microsoft.com. Physical attendance is capped at 2,500 -- the most intimate Build in years.

The three things to watch most closely:

• MAI coding model announcement: Reuters and The Information confirmed Microsoft will unveil homegrown AI models including a coding model to strengthen GitHub Copilot. If the MAI coding model has competitive benchmark performance against Claude Code and GitHub Copilot's current models, it directly addresses the billing backlash that went live today. The timing is not coincidental.
• Azure AI Foundry multi-model announcement: This formalizes Anthropic Claude and other non-OpenAI models as officially supported in Azure AI Foundry with enterprise SLAs. The competitive intelligence embedded in this announcement: Microsoft is signaling that the OpenAI-exclusive era of Copilot is over.
• Agent governance for enterprise: Agent 365, Entra Agent ID for identity management, and audit/compliance controls for autonomous AI agents in regulated environments. This is the governance product that will determine whether enterprises can actually deploy AI agents inside HIPAA, SOX, and GDPR-governed workflows at scale.

The GitHub Copilot token billing launched today will be the political backdrop for every developer-focused Build session tomorrow. Microsoft's messaging challenge: "We're making Copilot more powerful and more expensive today, but here is the cheaper, better model that makes it worth it."

For our full Build 2026 preview including all expected announcements and the Windows Agent Framework context, our AI News Today May 31, 2026 has the complete breakdown.

 Frequently Asked Questions

How much did Anthropic raise and what is its new valuation?

Bloomberg confirmed that Anthropic raised $65 billion in a funding round that values the company at $965 billion post-money, making it the most valuable private AI company in the world, surpassing OpenAI's $852 billion private market valuation. This supersedes the earlier $30 billion at $900 billion figures reported through mid-May; the final round size more than doubled. The round includes equity participation from Sequoia, Dragoneer, Altimeter, Greenoaks, Microsoft, NVIDIA, Founders Fund, and General Catalyst, as well as an associated $36 billion private credit facility from Apollo Global Management and Blackstone to purchase Google TPU chips for Anthropic's infrastructure.

What is the Apollo Blackstone $36 billion Google TPU deal?

Apollo Global Management and Blackstone are arranging a $36 billion private credit deal to purchase Google custom TPU (Tensor Processing Unit) chips on behalf of Anthropic, in what would be the largest private credit deal and largest chip-financing debt transaction in history. The debt is structured in three tranches (approximately $6B A1, $25B A2, and $4.5B B notes) through a special-purpose vehicle that buys the chips and leases them to Anthropic for use at data centers in New York, Texas, Louisiana, and Indiana. Broadcom, which co-develops the TPUs with Google, is backstopping payments on the largest tranches. Apollo and Blackstone are syndicating portions of the debt to outside investors while retaining significant portions themselves. Closing is expected as early as next week.

What is the SoftBank France AI investment?

SoftBank Group announced at the 2026 Choose France summit hosted by President Emmanuel Macron that it will invest up to €75 billion ($87.5 billion) to develop and operate 5 gigawatts of AI data center capacity in France. The first phase commits €45 billion to deliver 3.1 GW of capacity in the Hauts-de-France region by 2031, at three planned sites in Dunkirk (Loon-Plage), Bosquel, and Bouchain. SoftBank cited France's low-carbon electricity grid, industrial land availability, and engineering talent as primary factors. This is Europe's largest single announced AI infrastructure investment. The capital will come from a combination of fund raises, structured debt, and joint venture partnerships, as SoftBank's liquid balance sheet is approximately $30 billion.

Why is GitHub Copilot's new billing causing backlash?

GitHub Copilot switched from flat subscription pricing to token-based billing on June 1, 2026. Developers who previously paid $29/month for unlimited use are now seeing estimates suggesting their costs could balloon to $750/month or more if they use heavy agentic workloads, reasoning models, or large-codebase refactoring sessions. The phrase "What a joke" trended on Reddit and X on May 30 in anticipation of the change. Microsoft is announcing a new homegrown MAI coding model at Build 2026 on June 2 that is expected to be more cost-efficient than the current models underlying Copilot, which may alleviate the billing concerns for developers who switch to it.

Who is Cognition and what did they raise?

Cognition is a San Francisco AI startup founded in 2023 that makes Devin, an autonomous AI software engineer agent. On May 28, 2026, Cognition announced it raised more than $1 billion at a $25 billion pre-money ($26 billion post-money) valuation. The round was co-led by Lux Capital, General Catalyst, and 8VC, with participation from Founders Fund, Ribbit Capital, and others. Cognition's revenue grew from $37 million annualized in May 2025 to $492 million in May 2026, a 1,230% increase in 12 months. Enterprise customers include Goldman Sachs, Mercedes-Benz, NASA, and Santander. More than 90% of Cognition's own internal code is written by Devin.

What is CVE-2026-48710 BadHost?

CVE-2026-48710, labeled "BadHost," is a critical host header injection vulnerability in the Starlette Python web framework that allows unauthenticated remote attackers to bypass authentication by manipulating the HTTP Host header. The vulnerability affects FastAPI applications, vLLM deployments, LiteLLM instances, and every MCP server built on these frameworks, potentially imperiling millions of AI agents and AI-powered applications. Security firm Sysdig documented the first confirmed live cyberattack using an LLM agent that autonomously exploited this vulnerability (via a Marimo notebook) to exfiltrate an AWS database in under one hour without human direction of individual steps. Patches are available and organizations should update immediately.

What will Microsoft announce at Build 2026 on June 2?

Microsoft Build 2026 opens June 2 at Fort Mason Center in San Francisco with a free online livestream at build.microsoft.com. Based on Reuters and The Information reporting, confirmed announcements include: a suite of homegrown MAI AI models, including a coding model for GitHub Copilot (directly addressing today's billing backlash), and specialized models for transcription, reasoning, speech, and images. Also expected: Windows Agent Framework APIs for autonomous AI agents in the OS, Copilot Agent Mode for multi-step GitHub coding workflows, Azure AI Foundry multi-model updates formally adding Anthropic Claude, and Agent 365 enterprise governance capabilities. Satya Nadella delivers the opening keynote.

Recommended Reads

• AI News Today -- May 31, 2026: Claude Opus 4.8, Microsoft Goes Independent, and a Month That Changed Everything -- Build Fast with AI
• AI News Today -- May 29, 2026: The AI Cost Reckoning Has Arrived -- Build Fast with AI
• AI News Today -- May 22, 2026: Anthropic Goes Profitable, OpenAI Files for IPO -- Build Fast with AI
• Claude Mythos: Release Date, Access, and What Comes Next (2026) -- Build Fast with AI
• Claude Opus 4.8 Review: Benchmarks, Dynamic Workflows, and Price -- Build Fast with AI
• Best AI Models April 2026: Ranked by Benchmarks -- Build Fast with AI

References

• Bloomberg -- Apollo Shops $36 Billion Debt Deal to Buy Google Chips for Anthropic (May 28, 2026)
• Yahoo Finance / Bloomberg -- Cognition AI Coding Startup Raises $1 Billion at $26 Billion Value (May 28, 2026)
• The Next Web -- Cognition just raised $1 billion at a $26 billion valuation, and 90% of its own code is written by its AI
• TechCrunch -- SoftBank says it will invest up to €75 billion to build French data centers (May 30, 2026)
• CNBC -- SoftBank to build AI data centers in France with €75 billion investment (May 31, 2026)
• Data Center Dynamics -- SoftBank plans up to 5GW data center buildout in France
• TechCrunch -- "What a Joke": GitHub Copilot's new token-based billing spurs consternation among devs (May 30, 2026)
• AI Weekly Live Dashboard -- Foundation Phantom humanoid robots Ukraine, Sysdig LLM cyberattack, Wikipedia AI strike (June 1, 2026)
• TechTimes -- AI Regulation 2026: CNN Sues Perplexity, OpenAI Aligns With EU Rules (May 31, 2026)
• Crypto.news -- Blackstone and Apollo line up $36 billion chip debt deal for Anthropic
• Windows News AI -- Microsoft Build 2026: AI Agents, Copilot, Azure AI Foundry