Trending Thursday #42

In partnership with

Hi !

Here's what's been trending:

• OpenAI is raising up to $120B and killed Sora in the same week. Disney terminated its $1B deal. Safety moved under research.

• Meta cut 700 jobs while hiring AI agents. Two jury verdicts totaling $381M in damages. Reality Labs is reorganizing into AI pods.

• Arm shipped its first in-house chip. Stock jumped 10%. 136 cores, 2x performance per rack. Meta and OpenAI are early clients.

• Trump stacked his tech council with CEOs. Zuckerberg, Huang, Brin, Ellison, and Dell on a 13-member PCAST panel.

• Sanders and AOC want to freeze data centers. The moratorium bill would halt all new construction until federal safeguards are in place.

Startup Intros has a simple mission: to help tech founders navigate everything in the early-stage, from compliance to capital.

🔥 Startup Intros Events Coming Up

• Mar 30: AI Startup Pitch Night: AI founders pitch in front of an incredible group of investors. Unfortunately, pitches are closed, but register here!

• Apr 3: Caffeine & Capital: Our morning coffee meetup is back at Corgi Cafe in FiDi! Grab coffee, meet founders and investors & register on Luma!

• Apr 28-29: Startup Grind Conference: 5,000+ founders, investors & operators at the Fox Theatre in Redwood City, and we're planning a happy hour! Get $100 off with code STARTUPINTROSGC!

🎟️ Want a FREE ticket? Refer Us To Your Friends!

We just launched a referral program: Every person you refer to Startup Intros counts as one entry toward a free conference pass. Winner drawn March 31st!

🔵 OpenAI's $730B Reckoning

OpenAI is closing the largest private funding round in history this week, then killed its video product, restructured its leadership, and revamped its consumer experience.

💰 The $120B Round & the Org Shakeup

OpenAI CFO Sarah Friar confirmed the company is raising an additional $10B from a16z, D.E. Shaw, MGX, TPG, Coatue, and Thrive Capital, bringing the total fundraise to "north of $120B" at a $730B valuation. To put that number in perspective: OpenAI is now worth more than Walmart, JPMorgan, or Samsung. It's the most valuable private-to-public transition in tech history, and they haven't IPO'd yet.

The Dramatic Changes: Sam Altman told staff he has ceded oversight of safety and security to focus on fundraising, supply chains, and building data centers at scale. Fidji Simo was named CEO of a new division called "AGI Deployment," with Safety moving under Research and Security moving under Scaling. The next model, codenamed "Spud," has finished pretraining.

The Restructuring: OpenAI is no longer organizing around safety as a standalone function. It's organizing around deployment speed and infrastructure scale. Altman is becoming a data center builder, not a model reviewer.

🎬 The Sora Kill & the Product Pivot

Surprisingly, OpenAI announced it will discontinue all Sora products, including the consumer app, the developer API, and the video feature inside ChatGPT. Some Sora team members were surprised by the decision, which came just a day after OpenAI published a blog post about Sora's safety standards.

The Fallout: Disney terminated its three-year, $1B licensing agreement, signed in December 2025, that granted OpenAI access to 200+ Disney characters for Sora. That's a billion-dollar relationship, gone overnight.

The Pivot: ChatGPT's shopping experience got a full revamp, now letting users upload images, describe items, set budget criteria, and get visual product recommendations. The message is clear: video generation didn't work as a consumer product. Shopping, search, and task management will.

Translation: OpenAI is raising $120B, restructured its entire leadership around deployment speed, killed the product Disney paid $1B to license, and pivoted to shopping and agents, all in the same week. Is the velocity becoming a risk?

Hiring in 8 countries shouldn't require 8 different processes

This guide from Deel breaks down how to build one global hiring system. You’ll learn about assessment frameworks that scale, how to do headcount planning across regions, and even intake processes that work everywhere. As HR pros know, hiring in one country is hard enough. So let this free global hiring guide give you the tools you need to avoid global hiring headaches.

Download the free guide today

📱 Meta's Slow-Motion Restructuring

Meta is cutting, reorganizing, litigating, and hiring simultaneously. The company that spent $50B on AI infrastructure last quarter is still figuring it out.

✂️ The Cuts

Meta laid off approximately 700 employees this week, primarily in Reality Labs, as well as in recruiting, sales, and Facebook. The cuts follow the 15,000+ layoff plan, making this the first wave of what could be Meta's largest workforce reduction since 2023.

Reality Labs is reorganizing into AI-native "pods" focused on specific outcomes, flattening the leadership structure. Separately, Meta is offering top executives stock options for the first time since its 2012 IPO, tied to stock-price milestones over the coming years. The retention signal is loud: Meta understands its best people have options, and it's paying to keep them.

🤖 The AI Rebuild

While cutting in some areas, Meta is aggressively hiring in others. Meta acquired the Dreamer team, which builds AI agents, bringing on Hugo Barra (former Xiaomi and Meta VR VP), David Singleton (ex-Stripe CTO), and designer Nicholas Jitkoff. It seems that Meta is serious about consumer-facing AI agents, not just infrastructure.

CTO Andrew Bosworth was named to oversee "AI for Work," replacing Guy Rosen, with a mandate to drive AI adoption across Meta's entire workforce. And Meta launched Meta Small Business, a company-wide initiative led by Dina Powell McCormick to push AI tools to small business owners on Facebook and Instagram.

⚖️ The Legal Reckoning

The courtroom news was brutal. A New Mexico jury found Meta violated the state's unfair practices act by failing to safeguard its apps from child predators, awarding $375M in civil damages. Days later, an LA jury found Meta and YouTube negligent in the landmark social media addiction trial, awarding $6M in damages, with Meta paying 70% and YouTube 30%.

Also, Meta Ray-Ban Display's EU rollout has been hampered by regulations on AI features and removable batteries, as well as supply constraints. The hardware bet that was supposed to define Meta's next era is stuck at the border.

Translation: Meta laid off 700 people and hired an AI agent team while paying retention bonuses for the first time in 14 years and facing $381M in legal damages from two trials. Meta tries to play hard, but it just seems to be getting lost in the sauce.

💻 Arm's AGI CPU: The Chip That Could Change Everything

Arm, the company that designs chips for virtually every smartphone on earth, just launched its first in-house chip and bet its future on AI data centers.

🧠 The Chip

Arm launched the AGI CPU, its first in-house data center chip, featuring 136 Arm Neoverse V3 cores, 6GB/s of memory bandwidth per core, and more than 2x the performance per rack compared with x86 systems. It's fabricated by TSMC and designed specifically for agentic AI workloads in data centers.

Meta and OpenAI are early clients. Arm stock jumped 10%+ on the announcement. The chip positions Arm not as a licensor of designs but as a direct seller of silicon, competing head-to-head with Intel and AMD for data center market share.

📈 The Bet

CEO Rene Haas projected $25B in revenue by 2031, up from $4B+ in 2025, with approximately $15B expected to come from the in-house chip at a roughly 50% gross profit margin. That's a 6x revenue jump in five years, with the in-house chip accounting for 60% of the total.

The Risk: In a Wired Q&A, Haas acknowledged that Arm is now competing with its own licensees. Qualcomm, Samsung, MediaTek, and Apple all design chips on Arm architecture. Selling a competing chip could "piss everyone off," as Wired put it, but Haas argues the AI data center market is large enough that Arm's chip expands the pie rather than stealing slices.

The Context: SoftBank, which owns Arm, has committed $30B more to OpenAI and warned its loan-to-value ratio could exceed its 25% limit. Arm's in-house chip isn't just a product bet; it's the revenue engine SoftBank needs to fund its AI empire.

Translation: Arm went from designing chips for others to selling its own, overnight. So when you start competing with the companies that license your architecture, the 6x revenue projection comes with an asterisk if your biggest customers decide to build their own?

🏛️ AI Meets Washington

The White House and members of Congress announced two completely different visions for AI's future in the same week.

🇺🇸 Trump's Science and Tech Council

President Trump named 13 members to the President's Council of Advisors on Science and Technology (PCAST), co-chaired by David Sacks and long-time White House advisor Michael Kratsios. The 13-member roster reads like a who's who of AI infrastructure: Mark Zuckerberg, Jensen Huang, Larry Ellison, Sergey Brin, Michael Dell, and David Friedberg are all members.

The Panel: Advises on AI policy, research priorities, and technology regulation. Given the membership, the advice will skew heavily toward accelerating AI deployment rather than constraining it. Every major data center builder and chip supplier has a seat at the table, while consumer advocates, labor representatives, and safety researchers do not.

🛑 The Data Center Moratorium

On the other side of Pennsylvania Avenue, Sen. Bernie Sanders and Rep. Alexandria Ocasio-Cortez introduced the AI Data Center Moratorium Act, a bill that would pause all new data center construction across the US until federal safeguards for workers, consumers, the environment, and civil rights are in place. Axios warns the bill could theoretically delay new construction for years, depending on how the Senate interprets "strong" safeguards.

In a speech on Tuesday, Sanders argued: "A moratorium will give us the chance to figure out how to make sure that AI benefits the working families of this country, not just a handful of billionaires who want more and more wealth and more and more power."

⚖️ The Anthropic-Pentagon Standoff Deepens

A federal judge said the Pentagon's three actions toward Anthropic are "troubling" and that they look "like an attempt to cripple Anthropic." Separately, Anthropic filed a brief arguing it cannot manipulate Claude once the military has deployed it, denying DOW's accusations that Anthropic could tamper with models during wartime operations.

Translation: AI is causing various officials across our government to pursue different AI agendas, none of which are coordinated. It seems we are divided more than ever.

🚀 Product Launch Quick Hits

Anthropic ships Computer Use: Claude can now control your desktop directly. Available in research preview on macOS for Pro and Max subscribers.

Claude Code goes Autonomous: New auto mode makes permission-level decisions on its own while blocking destructive commands like mass file deletion.

ARC-AGI-3 redefines Benchmarking: Video-game-style scenarios measure on-the-fly reasoning instead of memorization. The old benchmarks are obsolete.

Apple reboots Siri from Scratch: Standalone app, overhauled interface, and full Gemini access in Apple data centers for distillation.

iOS 26.4 gets AI Playlists: Playlist Playground generates Apple Music mixes from text prompts. Eight new emojis and improved keyboard accuracy.

Beehiiv launches MCP Integration: Creators can now manage newsletter accounts through AI platforms with subscriber analysis and SEO optimization.

Google takes Android into Cars: Automotive OS expands beyond infotainment into non-safety internal vehicle systems. Software-defined cars are coming.

Ai2 open-sources Web Agent: MolmoWeb navigates browsers via screenshots instead of parsing HTML. Available in 4B and 8B parameter sizes.

Databricks enters Cybersecurity: New Lakewatch SIEM service already used by Adobe. Also acquired security startups Antimatter and SiftD.

Google Lyria 3 Pro drops: AI music generation jumps from 30-second clips to full three-minute tracks with better creative control.

Spotify fights AI Music Fraud: New beta tool lets artists review releases before they go live to prevent AI tracks from being attributed to them.

Reddit starts Labeling Bots: Automated accounts will be labeled going forward. Suspected bots must verify they are human to keep posting.

Google cCacks LLM Compression: TurboQuant quantization algorithm enables massive model compression without sacrificing accuracy. Open infrastructure play.

Kalshi bans Insider Trading Preemptively: Politicians, athletes, coaches, and officials are blocked from trading in politics and sports prediction markets.

👔 Personnel Quick Hits

Jensen Huang declares "We've Achieved AGI": Told Lex Fridman in a wide-ranging interview covering Nvidia, scaling laws, OpenClaw, China, and data centers in space.

OpenAI poaches JioStar CEO for APAC: Kiran Mani takes a newly created Asia-Pacific role reporting to CSO Jason Kwon. OpenAI's international push accelerates.

Microsoft raids AI2 for Suleyman's Team: Former CEO Ali Farhadi and key Allen Institute researchers join Mustafa Suleyman's AI team in Seattle.

Epic Games axes 1,000+ Employees: Fortnite maker admits it's "spending significantly more" than it earns. Targeting $500M in projected savings.

Baltimore sues xAI over Grok: City accuses Musk's AI company of violating consumer protection laws and engaging in deceptive trade practices.

China traps Manus Co-founders In-country: Beijing reviewing whether Meta's $2B acquisition violates foreign direct investment rules. Co-founders barred from leaving.

🌟 Editor's Note

At Startup Intros, our mission is to bring the latest founder-investor news straight to your inbox, keeping you ahead in the fast-paced world of Silicon Valley.

💭 Parting Thoughts

OpenAI is raising $120B at a $730B valuation while killing Sora and moving safety under research. Meta laid off 700, hired an AI agent team, and got hit with $381M in damages from two trials. Arm shipped a chip that sent its stock up 10% and could generate $15B in annual revenue.

The White House stacked its AI council with CEOs. Bernie wants to freeze AI data center construction. A federal judge questioned whether the Pentagon is trying to destroy Anthropic.

There’s a widening gap between what's being built and what's being governed every week. Till next time!

Startup Intros Trending Thursday: Your trusted source for founder-investor insights, delivered with clarity and focus.