April 10th Newsletter

Author

Da Yu
April 10, 2026

As of close on April 9th, 2026

Markets are navigating a volatile mix of sticky inflation, geopolitics, and corporate restructuring. Core PCE inflation held at 3% (headline 2.8%), above the Federal Reserve target, with the Fed's March minutes showing a willingness for rate hikes. Oil markets swung sharply, with WTI down 16% to $93.98 on Tuesday after President Trump announced a two-week Iran ceasefire, though uncertainty persists as reports emerged later that day that the Strait of Hormuz was closed after a brief reopening due to Israel continuing their campaign in Lebanon. Cyber risk also escalated as the FBI, NSA, DOE, and CISA warned of Iran-linked attacks on US critical infrastructure. In tech and logistics, Amazon struck a new deal with the United States Postal Service to stabilize roughly $6B in postal revenue. In the media, Disney is planning up to 1,000 layoffs, mainly in marketing, amid consolidation under new CEO Josh D’Amaro. Lastly, Oracle appointed Hilary Maxson, former Schneider Electric CFO, as CFO to manage AI-driven cloud expansion amid rising debt.

Top 5 AI Highlights

[1] 🔒 Anthropic launched a new cybersecurity initiative called Project Glasswing, giving limited access to its new Claude Mythos Preview model to select partners to safely test advanced vulnerability-detection capabilities while reducing risks of misuse in cyberattacks. Additionally, Apple, Microsoft, Google, Amazon, and Nvidia joined to form a 12-company alliance, backed by $4 million in direct cash donations to open-source security groups and $100M in Claude usage credits. More from Anthropic and ZDNet.

[2] 💰 Anthropic has committed to multiple gigawatts of Google/Broadcom TPU capacity starting in 2027 as run-rate revenue topped $30B, up ~233% from $9B at end-2025, with now over 1,000 customers spending more than $1M, compared to just 500 as reported in Feb 2026 during the Series G announcement. Separately, Anthropic is planning a $200M commitment to a $1B PE-backed venture selling Claude adoption services. Details on Anthropic and WSJ.

[3] ✨ Meta launched “Muse Spark,” an AI model marking a major overhaul under its new Superintelligence Labs; it introduces multi-agent reasoning, a planned “Contemplating” mode, and aims to improve complex problem-solving. Also, CoreWeave signed a $21B expanded AI infrastructure deal with Meta through 2032. More from Meta and Coreweave.

[4] 🏭 Intel joined Elon Musk’s Terafab project in Texas with Tesla and SpaceX to “design, fabricate, and package ultra-high-performance chips at scale” for advances in AI and robotics. The company’s share price rose 3%. Details from Intel and TechCrunch.

[5] 🚀 SpaceX is strongly encouraging IPO advisers to purchase Grok subscriptions, with banks spending millions, underscoring Musk’s leverage as Wall Street chases a $50 billion offering potentially valuing SpaceX over $1.75 trillion. More from NYTimes.

✨Open Source AI Development

[1] 🦞 Anthropic will charge Claude Code subscribers separately for OpenClaw and other third-party harnesses, citing sustainability and engineering limits, intensifying tensions with open-source developers as OpenAI backs OpenClaw. Read more on TechCrunch.

[2] 💼 Z.ai launched an open-source model, GLM-5.1 (754B MoE), enabling 8-hour autonomous agentic coding. It reportedly beats GPT-5.4 and Claude Opus on SWE-Bench Pro, boosting enterprise AI competition. Details are on VentureBeat.

[3] 🤝 OpenAI, Anthropic, and Google are sharing data via the 2023 Frontier Model Forum to curb Chinese model copying, with the potential to protect billions in profit and hundreds of billions in AI infrastructure spending. Info is on Bloomberg.

[4] 🐴 Alibaba reportedly anonymously released “HappyHorse-1.0,” a leading open-source AI video model that topped leaderboards and surpassed ByteDance’s Seedance 2.0 amid intensifying AI video competition and stealth product launches in China. See the details on The Information.

[5] 🇺🇸 Arcee, a 26-person U.S.-based startup, released the Trinity Large Thinking 400B-parameter open-weight reasoning model built on a $20M budget, targeting Western enterprises seeking alternatives to Chinese models with performance that’s comparable to models from Anthropic, Kimi, and Z.ai. Learn more about it on TechCrunch.

Founder’s Corner

Strong Start to 2026 for Hardware

Source: X

What We Read This Week

  • Amazon CEO Andy Jassy defended the company’s $200B capex, citing strong AI demand: AWS AI has a $15B run rate, chips are generating $20B+ annually, and despite FCF falling, the company has long-term growth confidence. Share prices rose 4.5% on the news. (GeekWire)

  • OpenAI paused its Stargate UK data center project over high energy costs and spending restraint ahead of IPO, while continuing US, Norway, and UAE builds amid competition and regulatory negotiations. (Bloomberg)

  • TikTok will invest €1B in a second data centre in Lahti, Finland, expanding European data sovereignty (50MW initial, 128MW max) by 2027, serving 200M users amid EU scrutiny. (Reuters)

  • A new report from the New Yorker details the chaos surrounding Sam Altman’s firing – and return – at OpenAI, his political relationships, and questions regarding his commitments to AI safety. (The New Yorker)

  • Elon Musk amended his lawsuit against OpenAI, seeking damages to be awarded to its nonprofit arm and pushing for Sam Altman’s removal from the board, escalating tensions over OpenAI’s shift toward a for-profit structure. (WSJ)

  • Google’s AI Overviews are ~85–91% accurate based on recent tests, but often include ungrounded or misleading information; they draw from mixed-quality sources, including social media platforms like Facebook and Reddit, making fact-checking necessary despite their authoritative appearance. (New York Times)

  • A report from the New York Times detailed how Medvi, a GLP-1 telehealth provider, scaled from launch to a projected $1.8 billion revenue run rate with two employees using AI-heavy operations. However, the report is now receiving backlash after investigations alleged that Medvi used fake patient images, AI-generated testimonials, and deceptive affiliate ads. The company also received an FDA warning. (Read the original New York Times report and look through a concise summary of the allegations on Morning Brew.)

  • Kalshi struck a wide-ranging Fox deal to embed prediction forecasts across Fox News, Fox Business, Fox One, and Fox Weather, expanding beyond existing CNN and CNBC partnerships; election coverage is excluded. (The Hollywood Reporter)

  • Pershing Square proposed merging Universal Music Group into its SPARC vehicle, offering €5.05 plus 0.77 shares per share, implying a 78% premium and potential U.S. relisting. (Bloomberg)

AI Fundraising News (Apr 3 — Apr 9)

  • AfterQuery: AI expert-data platform company, raised a $30M Series A

  • Alien: AI human verification network company, raised a $7.1M pre-seed round

  • Anvil Robotics: AI physical robotics platform company, raised a $5.5M seed round

  • Aria Networks: AI networking infrastructure company, raised a $125M Series A

  • Astromech: AI biological data company, raised a $10.5M seed extension

  • Chapter: AI Medicare enrollment platform company, raised a $100M Series E

  • CONXAI: AI construction workflow automation company, raised a $5.8M Series A

  • Elorian: AI visual reasoning models company, raised $55M

  • Firmus: AI data center infrastructure company, raised $505M

  • Generare: AI microbial drug discovery company, raised a $22.9M Series A round

  • Handhold: AI sales agents company, raised a $3M seed round

  • Hermeus: AI defense aircraft company, raised a $350M Series C

  • Mappedin: AI indoor mapping company, raised $24.5M

  • Miravoice: AI voice agent company, raised a $6.3M seed

  • Modus: AI audit workflow company, raised an $85M Seed & Series A

  • Moonbounce: AI governance controls company, raised $12M

  • Natter: AI enterprise insights company, raised a $23M Series A

  • Nava: AI full-stack neo-cloud infrastructure company, raised a $22M Series A

  • NeuBird AI: AI incident response company, raised $19.3M

  • Noon: AI product design company, raised a $44M seed

  • Numos: AI financial data analysis company, raised a $4.25M seed round

  • Omniscient: AI executive monitoring agents company, raised a $4.1M pre-seed round

  • Patlytics: AI patent law platform company, raised a $40M Series B

  • Plume: AI renewable geospatial platform company, raised $3.3M

  • Poke: AI text-message agent platform company, raised $10M

  • Ridge AI: AI embedded analytics company, raised a $2.6M pre-seed

  • Rork: AI mobile app development company, raised a $15M seed round

  • Sarvam: AI Indian language voice company, raised $350M

  • Satellites on Fire: AI wildfire detection platform company, raised a $2.7M seed round

  • SiFive: AI RISC-V data center chip design company, raised $400M

  • Tenex: AI-native cybersecurity monitoring company, raised a $250M Series B

  • Trent AI: AI security agents company, raised a $13M seed

  • True Footage: AI property valuation company, raised a $40M Series C

  • Xoople: AI satellite mapping company, raised a $130M Series B