OpenAI's $840B Flex: Double the Headcount, Double the Drama

Spotted: Sam Altman quietly plotting world domination from a WeWork in SOMA. After raising a casual $110 billion at an $840 billion valuation, OpenAI announced plans to nearly double its workforce from 4,500 to 8,000 employees by end of 2026. The irony? This hiring frenzy comes just two months after Altman's January town hall where he claimed AI automation would let them "get vastly more done with far fewer people." Turns out, even AI needs human "technical ambassadors" to close those billion-dollar enterprise deals. The plot twist nobody saw coming — except maybe Anthropic, who's already captured 40% of the enterprise AI market while OpenAI dipped to 27%. [CNBC] [Lapaas Voice]

Oasis Security Banks $120M for the Agent Apocalypse

While you were sleeping, Oasis Security raised a $120M Series B (total: $195M) to solve a problem most people haven't heard of yet: managing access for AI agents and "non-human identities." Because apparently, when the robots take over, they'll need enterprise SSO too. The NY-based startup is betting that as AI agents proliferate, someone needs to be the bouncer at the enterprise door. Smart money says they're not wrong — just early. [SecurityWeek]

Frore Systems: The Cool Kid with $143M at $1.6B

In a world overheating with AI data centers, Frore Systems just raised $143M in Series D funding at a $1.6B valuation to keep things chill — literally. The San Jose-based startup makes cooling systems for data centers, and with energy demand from AI infrastructure exploding, their timing couldn't be more perfect. When your customers are burning through terawatts, being the air conditioning guy is suddenly very good business. [Axios Pro Rata]

The Technical Ambassadors Are Coming

OpenAI isn't just hiring engineers — they're creating a whole new job category: "Technical Ambassadors." These aren't your average sales engineers, darlings. Think of them as AI whisperers tasked with helping businesses actually use OpenAI's tools instead of just paying for them and hoping for the best. It's a smart move in a market where 70% of first-time enterprise AI buyers are choosing Anthropic over OpenAI. When the product isn't selling itself, hire people to sell it. Revolutionary. [Reuters]

XBOW Hits Unicorn Status with $120M at $1B+

In today's episode of "AI makes everything scarier," XBOW raised $120M at a $1B+ valuation for its autonomous offensive security platform. The startup uses AI to automatically discover and validate software vulnerabilities — essentially letting the machines hack themselves before the bad guys do. It's like having a white-hat hacker who never sleeps, never eats, and definitely never asks for equity. [SecurityWeek]

Native Exits Stealth with $42M to Kill Cloud Drift

Native emerged from stealth this week with $42 million in funding to solve "cloud drift" — the phenomenon where your cloud infrastructure slowly morphs into something unrecognizable while you weren't looking. Think of it as Marie Kondo for your AWS bill, except instead of asking if things spark joy, it just silently fixes your infrastructure. In an era where abstraction is rising and compute is tightening, operating smarter is becoming the new competitive advantage. [Startup Club]

Turquoise Health's $40M Prescription for Healthcare Billing

Healthcare pricing startup Turquoise Health raised $40M in Series C funding led by Oak HC/FT with a16z participating, bringing total funding to ~$100M. The company evolved from a pricing data vendor to a full workflow platform using AI to digitize payer contracts and establish single, auditable rates. In a world where a band-aid can cost $800, someone's finally bringing spreadsheets to the healthcare price transparency fight. [Fierce Healthcare]

Jensen Huang's "Token Economics" Has Everyone Talking

The leather-jacket-wearing oracle of Nvidia, Jensen Huang, dropped a bombshell at GTC this week: he wants to pay engineers in AI tokens. His vision? Engineers get a token budget equal to half their salary to deploy AI agents as "productivity multipliers." As Huang put it: "If that $500,000 engineer did not consume at least $250,000 worth of tokens, I am going to be deeply alarmed." Tomasz Tunguz of Theory Ventures has been preaching this since February, calling inference costs the "fourth component of engineering compensation." Is this the future of work or just a very expensive way to make sure your employees actually use the tools you pay for? Time will tell, darlings. [TechCrunch] [Business Insider]

Garry Tan on What Founders Should Actually Build

Y Combinator's President Garry Tan had some advice for agentic systems founders this week: "People do not want to pay for raw markdown and they shouldn't have to. But they may pay for orchestration, hosting, updates, collaboration, portability, analytics, and managed execution." Translation: Don't just build the AI wrapper — build the whole damn sandwich. Also spotted: Tan reminding everyone that YC "probably funds more [non-tech-heavy startups] than all of Silicon Valley combined" and turns them into "IPO-class businesses." The man's not shy about YC's track record, and honestly, why should he be? [X/Twitter] [X/Twitter]

Delve's "Fake Compliance" Scandal Explodes

In today's episode of "Security Theater Gone Wrong," compliance startup Delve is facing accusations of "fake compliance" from customer DeepDelver. The allegations? That Delve "achieves its claim of being the fastest platform by producing fake evidence, generating auditor conclusions on behalf of certification mills that rubber stamp reports, and skipping major framework requirements while telling clients they have achieved 100% compliance." Even worse, Delve allegedly helps customers "mislead the public by hosting trust pages that contain security measures that were never implemented." It's one thing to fake it till you make it — it's another to fake security compliance. That's how people get hacked, sued, and very publicly shamed on TechCrunch. [TechCrunch]

Elon's Terafab: The Chip Dream That May Never Happen

Never one to let a week pass without announcing something that sounds impossible, Elon Musk unveiled "Terafab" — a joint Tesla, SpaceX, and xAI project to manufacture chips in Austin. The goal? Produce a terawatt of compute annually for everything from Optimus robots to "space-based AI satellites." The catch? Musk has "no background in semiconductor production and a history of over-promising on goals and timelines." Also, he's still ordering Nvidia chips "at scale" because apparently you can't just manifest a chip fab into existence. But hey, when you've already promised robotaxis, Mars colonies, and fully autonomous Taxis, what's one more moonshot? [Bloomberg] [Business Insider]

Nvidia's $1 Trillion Chip Forecast

Jensen Huang isn't known for modest predictions, but this one's a doozy: Nvidia expects $1 trillion in orders for Blackwell and Vera Rubin chips through 2027. That's double the $500B demand he cited just a year ago. The market may have shrugged at GTC (Nvidia stock actually dropped during the keynote), but when someone projects a trillion dollars in chip sales, you listen. The Vera Rubin chips are designed for AI inference — the "thinking" part of AI — which is becoming the new bottleneck as models get smarter. [CNBC]

The European Funding Frenzy

While Silicon Valley obsesses over AI tokens, Europe is quietly having a moment. Montis VC reached a €50M first close for energy transition and AI startups. Parallel (Paris) raised $20M Series A for AI hospital billing agents. Rivia (Zurich) bagged €13M for clinical trial AI. Reson8 (Amsterdam) took €5M pre-seed for speech AI. The pattern? AI agents for specific verticals with European regulatory compliance built-in. Smart money is betting that GDPR-compliant AI will be a competitive moat, not just a headache. [The Next Web]

Meta's $80B Metaverse + 16K Layoffs = Cognitive Dissonance

In one of the more confusing corporate strategies of 2026, Meta is simultaneously laying off up to 16,000 employees (20% of workforce) while spending $80 billion on the metaverse. Zuck giveth, Zuck taketh away. The company also reversed its plan to kill VR support for Horizon Worlds after mobile downloads surged 53%. It's the tech equivalent of "we're not breaking up, we're just taking a break" — except with billions of dollars and thousands of careers. [Bay Area Letters] [LatestLY]

A little birdie (aka one of you gorgeous subscribers) sent this along: "Chew The Cud, Cognitive Dissonance is a Superpower" — a piece on embracing contradictory thoughts in the face of increasingly polarized discourse. In a world where everyone wants you to pick a side, maybe holding two opposing ideas at once is the real flex. [Read it here]

Got something to share? Hit reply and spill the tea. Anonymity guaranteed, darlings.