The OpenAI / Hugging Face hack stirred up discourse about the dangers of rogue AI, while the model-makers move to bring in big cybersecurity bucks. Dimension founder Zavain Dar explains what Silicon Valley can learn from Chinese AI companies on the podcast. Inference provider Baseten scoops up an early-stage infrastructure startup. Mistral brings in billions but can’t shake its identity crisis. A new Ramp report shows AI spending per employee dipped in August. Coding assistant Cognition rakes in billions. Meta launches its Instinct-competitor Muse. The DOJ turns its attention to Nvidia’s Groq acquihire. Biden administration officials tell their side of the infamous a16z meeting in 2024 that led firm leaders to endorse President Trump.

OpenAI & Anthropic Could Profit from Cyber-Defense Tools. The Irony Is Lost on No One.

The uproar over AI safety that dominated the tech conversation this week was something to behold: whistleblowers declaring the end is near, politicians in a panic about how to handle what will likely be the most consequential issue of their careers, and many in the industry deeply uncertain about the level of risk that further model improvements might bring.

The fears of an individual young researcher who worked at Anthropic for less than two months have gotten far more airtime than they deserve.

Whatever your level of p(doom), though, for anyone not directly involved in frontier model development or high-level policy, there’s a simple way to think about all this: it’s a cybersecurity problem.

That’s what was on the minds of founders we spoke with this week.

The Anthropic threat intelligence report published Thursday detailing bad actors’ efforts to use Claude in nefarious ways also underscores the point. Threat intelligence has been a tech growth industry for decades.

Top-tier AI models are very, very good at finding and patching — or exploiting — vulnerabilities in computer systems.

Cyber-defense is thus set to become a major line of business for OpenAI and Anthropic. It’s coming at a crucial moment for the model providers, which need to sustain vertiginous revenue growth ahead of their IPOs but are starting to see signs that established categories like coding may not be enough.

However you might feel about buying solutions from the people who caused the problem, you might not have much choice.

“Where there’s a vulnerability in a system you can now hack something in a few hours,” said Erik Bernhardsson, the co-founder of Modal. “We’ve been running those models internally; they’re very good at finding things.”

Bernhardsson said his AI infrastructure company has been using it as a partial replacement for expensive external security consultants. Other founders concur, and say they’ve been impressed with the capabilities of the frontier models for cybersecurity related tasks.

Jean-Denis Greze, co-founder of the buzzy AI assistant startup Town, said continuous security monitoring used to be too expensive to bother with. “Before it wasn’t worth it — now it is.”

AI Cybersecurity as a Revenue Opportunity

Another founder said that many of the vulnerabilities he’s seen these tools patching up come from older systems. That raises the possibility that we could see a spike in AI cyber-defense spending as people work to clean up their legacy (human-coded) infrastructure, but once those vulnerabilities are discovered and updated the need for these tools would decline.

On the other hand, a recent story in The Information highlighted a broader shift of cybersecurity budgeting toward AI tools and away from older software made by the likes of Cisco and SentinelOne. The Information credited the spring announcement of Anthropic’s Mythos model with setting off the scramble for tools to fight bad actors.

OpenAI certainly seems to be betting on long-term growth for the category. Last week, on the day it rolled out Astra, it held a separate cybersecurity session led by president Greg Brockman detailing the cyber capabilities of the new model. Sam Altman is out selling it. Even assuming that OpenAI ringing the alarm about Hugging Face arose from genuine concern, there’s no doubt that the current panicky climate around AI risk is a good one for cyber-defense tools.

Nvidia’s Jensen Huang even touted it as the next big AI thing at a Goldman Sachs conference in San Francisco on Thursday.

It’s not clear what kind of revenue opportunity cyber might represent for the big labs. So far, they’ve been riding the blowout success of agentic coding and the huge token bills companies are racking up. OpenAI has crossed $40 billion in annual run rate; Anthropic reportedly has crossed $65 billion.

But at least some of its biggest customers have grown more cost-conscious in recent months, new data from Ramp shows (see below).

Uber provides a good example of the shifting dynamic. Earlier this year the company revealed that it had blown through its entire AI budget several months into the year by spending on tokens. Last month it revealed its new approach in a blog, which showed that while its agent usage continued to go up, its cost per 1,000 model requests is down almost 34% from its peak. Essentially, usage was up significantly and spending was still increasing but spending growth has slowed.

We can imagine there are some tense conversations these days between the safety-conscious and the financially ambitious inside Anthropic and OpenAI. Cybersecurity as a focus is something they can agree on.

Newcomer Podcast

Zavain Dar on Hugging Face, Nvidia & Why Silicon Valley Should Pay Attention to China

European Models

Despite a Big Raise, Mistral May Not Shake Its Also-Ran Reputation

This week, the French AI company Mistral raised a €3 billion Series D at a €21 billion post-money valuation. The raise is the latest move from a company that has left observers bewildered by its strategy. PSG Equity, Samsung, and the EQT-managed Scaleup Europe Fund co-led the round. Heavy hitters that have followed on from previous rounds include a16z, Index Ventures, General Catalyst, Lightspeed Venture Partners, and Nvidia.

Mistral’s recent moves suggest that it is giving up on being “Europe’s answer to OpenAI” and pivoting to institutional deployments.

In July, it was reported that Microsoft was funding a multi-billion-dollar buildout of European compute for Mistral. Then, in August, Mistral announced it would begin hosting third-party open-source models on its platform, starting with Chinese firm Z.ai’s GLM-5.2.

Some outlets have interpreted these moves as Mistral following the Palantir playbook, using forward-deployed engineers to bring customers onto Mistral’s proprietary stack. However, one critic cited in Le Monde instead compared them to Capgemini, the far less glamorous IT solutions company.

A VC we also spoke with, who is familiar with the company, said, “The ambition has morphed from competing with OpenAI to competing with the Chinese models to implement models as European Palantir. The danger, however, is that you end up as that other French services company, Capgemini.”

This comparison is apt if customers prefer to use open-source models other than Mistral’s while still relying on the firm for AI services. Mistral’s “frontier-class” model, Medium 3.5, is the world’s 31st most powerful open-weight model per the capabilities index run by research firm Epoch AI, although it remains first among European firms.

Another question is whether the new French administration will continue to hold up Mistral as a continental champion. Politico highlighted this week that much of the firm’s success has been enabled by its close relationship with French president Emmanuel Macron, whose replacement in 2027 may well be the Eurosceptic Marine Le Pen.

Despite these doubts, Mistral says it intends to fight at the frontier. Global Affairs SVP Audrey Herblin-Stoop told Le Monde that the raise will fund a compute buildout that enables more powerful models and allows Mistral to host their customers’ AI systems, raising revenues to be invested back into research. Mistral CEO Arthur Mensch is well-respected in the industry, and if the shock success of Moonshot AI’s Kimi K3 in July is anything to go by, labs are only as good or bad as their last model.

AI Acquisitions

Baseten Deal Shows Appetite for Infrastructure Startups

Inference providers are looking for more ways to build out their tools to help run AI agents, a new deal shows.

On Thursday, Baseten announced that it had acquired the startup Blaxel, which has been developing sandboxes for AI agents running code or other background tasks. It’s the second acquisition for the inference provider in the last year, after purchasing the post-training startup Parsed in December 2025.

The price wasn’t disclosed in the press release, but we heard the deal total was roughly $300 million.

For Blaxel, it’s a strong markup — the startup had only raised $7.3 million in seed funding last December from First Round Capital, Liquid 2 Ventures, and Y Combinator.

Some investors we’ve spoken with in the past have hesitated to back many of these kinds of startups, expecting that the companies that needed the technology would build it themselves. But this deal shows at least one well-capitalized startup would rather go shopping.

Three Big Charts

Power Users Ease Off On AI Spending as Token Costs Fall

Ramp’s latest AI Index Report shows some signals that the extraordinary growth of AI spending may be starting to ease, though Ramp research head Ara Kharazian is cautious about drawing broad conclusions.

Read more

Alex Heath — a top tech journalist known for his interviews with tech heavyweights and for his consumer technology scoops — is joining Guy Oseary’s Sound Ventures as a partner.

He’s signing on to the Los Angeles-based firm as it’s targeting a $300 million fifth fund with an expected first close this month.

“I am hanging up my capital J journalist hat,” Heath told me in a conversation Wednesday. “That’s for sure happening.”

Still, Heath says he will continue running his newsletter and podcast, both called Sources, independently from Sound Ventures. Heath tells me the publication has generated more than $1 million in revenue (largely from sponsors) since he left The Verge and launched it in September 2025.

“This world is pretty blurry,” he says about his new status as a full-time VC and now a not-so-independent tech media persona. “There’s room for a lot of great conversations and insights and getting founders to open up — which I think I’m pretty good at.”

He’s making the transition on a high note having just published interviews with Mark Zuckerberg and Sam Altman and having profiled Altman in a cover story for Time magazine.

Subscribe now

Heath says he’s getting out of the adversarial scoop game but plans to continue publishing big interviews and writing about interesting startups (some of which he’ll invest in). He’ll be investing in startups for the first time ever in the new gig.

Sound Ventures was founded by Oseary and actor-turned-investor Ashton Kutcher more than a decade ago. The Wall Street Journal reported in June that Kutcher is leaving Sound to start another VC firm called Decimal Capital with Morgan Beller, a former general partner at NFX.

Kutcher left the firm after a string of major victories. The firm cut a $55 million check in OpenAI at a $27 billion valuation and a $35 million check in Anthropic at a $5 billion valuation in 2023 — way before large multi-stage firms were trying to invest in both AI juggernauts. Sound leveraged those early investments to write $370 million in SPVs into OpenAI and about $400 million in Anthropic. The firm was also an investor in Hugging Face and Altruist, which are selling for $12.9 billion and about $4 billion respectively.

Now Sound is out raising its fifth flagship fund. The firm is targeting $300 million, the same size as its previous fund, sources tell me, and expects its first close this month. Sound will continue to double-down on its more promising investments with special purpose vehicles. The firm has nearly $2 billion in assets under management and the firm values its holdings at $6.5 billion.

Heath is joining Oseary, the firm’s founder, and managing partner Effie Epstein as the only other person at the firm with a partner title. They’re all based in Los Angeles. (The firm recently hired Nicole Ripka as an investor and Jill Puente as its chief marketing officer.)

When I asked Oseary how he felt about Heath keeping Sources going as an independent company, he said, “I still manage Madonna. I still manage the Red Hot Chili Peppers. We still do a lot of things. We’re out there. We’re in the mix.”

Read more

We’re three weeks out from the Machine Earning AI Summit, bringing together the leading founders, investors, and operators focused on AI finance and commerce.

There’s still a bit of room left for top founders and investors to attend. If you’re interested in coming, you can apply on our site here.

Apply to Attend

We’re sharing the names of two additional speakers alongside our first set of discussion group leaders at our one-day summit.

Our lineup will now include on-stage panels with Ryan King, the co-founder of Chime, and Jean-Denis Greze, the co-founder of the buzzy personal assistant Town and former CTO of Plaid.

Additionally, we’re thrilled to announce our first set of discussion group leaders. Our breakout sessions put attendees into small, curated conversations with AI founders, investors, and operators, and they’ve become a highlight of our summits.

Our first batch of discussion group leaders includes:

• Peter Hazlehurst, founder & CEO of Synctera

• Erica Dorfman, CFO of Brex

• Kahlil Lalji, founder & CEO of Natural

• Darragh Buckley, founder & CEO of Increase

• Mackenzie Burnett, founder & CEO of Ambrook

• Nico Simko, founder & CEO of Clair

• Anu Maheshwari, head of venture ecosystem at Nebius

• Aditi Maliwal, partner at Upfront Ventures

Our upcoming summit will explore how AI is fundamentally changing finance and commerce. Sessions cover agentic commerce, intelligent money, enterprise finance operations and the shopping experience, plus the fraud, risk, and compliance questions running through all of them.

Many thanks to our sponsors Airwallex, Exa, Lead Bank, Nebius, and OnePay.

The summit will be held on Tuesday, September 29, in San Francisco. We can’t wait to see you there.

Tim Cook’s retirement at Apple highlights how few unifying leaders there are in Silicon Valley today. Tech CEOs call for lax AI regulations from international dignitaries at the G20 Innovation Ministerial. Too many AI auditors come from a similar POV. Anthropic and OpenAI’s professional power-users account for most of their enterprise revenue. Nvidia and Hugging Face announce the chipmaker’s $12.9 billion acquisition agreement. OpenAI pulls back the curtain on its new Astra model. Pablo Torre’s investigation into the LA Clippers results in historic NBA sanctions and a Pulitzer for his team. The Trump administration weighs in on behalf of OpenAI in its legal battle with the New York Times. SpaceX investor 137 Ventures shows strong returns after the record-shattering IPO.

The Main Item

There’s a Leadership Vacuum in Silicon Valley

With Tim Cook’s retirement, the tech industry has lost a leader who can fairly lay claim to being one of the great CEOs of all time. Certainly, by the standards that such things are currently judged, his record is unassailable: Apple had a market cap of $347 billion when he took over in 2011, and it now stands at nearly $4.7 trillion, a tidy 1200% return for shareholders in what was already a very, very large company.

Yet Cook’s departure underscores a troubling fact about the tech business today: it lacks leaders who command broad respect across the industry, just at the moment when it needs them most.

There is no Jamie Dimon in Silicon Valley, no Bob Iger, no Warren Buffett — businesspeople who know how to make money but also have a broad lens and an interest in what their decisions mean for the world. Eric Schmidt once played that role as CEO of Google. Back in the day there was Andy Grove at Intel, and before that Bob Noyce, brilliant innovators who were also dialed into the societal stakes of their work.

Instead, today, we have the likes of Mark Zuckerberg, Elon Musk, and Jeff Bezos — self-absorbed men who are relentlessly focused on their personal priorities and financial interests to the exclusion of everything else.

Even Cook, though he took principled stands on privacy and famously faced down the government over access to a mass-murderer’s iPhone, too often put profit over principle and shrugged off any leadership responsibilities beyond running Apple.

Like almost all of his compatriots, Cook suddenly went mute on his support for immigrants when the second election of Donald Trump led to ICE rampages in Minneapolis and elsewhere. He gave the President a literal golden gift, staining himself and his company with the administration’s extreme corruption. Apple was among the few that had the muscle to stand up to Trump’s cruel and often extra-legal adventures, but Cook chose instead to bend the knee and protect the bottom line.

In recent weeks, amid tense discussions of the AI risks exposed in the OpenAI/ Hugging Face hack and the sudden nationwide revolt against data centers, it would have been great to see top industry execs engaging with the issues and pushing one another to come up with real solutions.

Instead, it was left to Bill Gates, long since retired from Microsoft and tarred by his Jeffrey Epstein associations, to issue a rallying cry for collaborative industry action. It didn’t exactly send everyone scrambling.

Meanwhile, two of the most prominent figures in the tech startup ecosystem, Paul Graham and Palmer Luckey, spent their week hurling vicious insults at one another in a flame war over immigration. Dario Amodei and Sam Altman, the leaders of the two top AI companies, won’t even pretend to shake hands. Jensen Huang is busy being a rock star. Trump-aligned investors like Marc Andreessen and David Sacks seem more interested in owning the libs than engaging on difficult questions. It’s not very edifying.

Fifteen years ago, after Steve Jobs died, the conventional wisdom was that Cook’s existential challenge was to come up with a killer product of his own. The new CEO had to prove that he could be an innovator like Jobs if he was to take Apple to the next level; the company wouldn’t be able to prosper forever with just the iPhone.

That turned out to be an all-time bad take. Cook can claim the Apple Watch, but his real success was in optimizing Jobs’ inventions. That turned out to be more than enough.

Apple’s new CEO, John Ternus, is now in the unenviable position of being judged on both metrics: can he innovate like Jobs, and run the ship like Cook?

Ternus is a low-profile hardware engineer and the public-facing aspects of the job don’t obviously play to his strengths. But like Cook, he has an opportunity to succeed in unexpected ways. In the new world order, technical chops plus the ability to play politics and public opinion could prove the key to success. The position of wise and widely respected industry leader remains wide open.

Machine Earning

Apply Today to Attend the Machine Earning AI Summit on Sept. 29 in San Francisco

We’re less than a month away from the Machine Earning AI Summit. Spots are limited. Apply today for a chance to attend.

Apply Here

Yesterday, we published our market map of companies that are working on agentic shopping, smart payments, and the next-gen infrastructure rails that will power transactions.

AI Policy

Tech Leaders Push Trump’s AI Agenda at G20

Anthropic and the Trump administration publicly squashed their beef at this week’s G20 meetings — one in Asheville, North Carolina, focused on finance and another in Chapel Hill dedicated to technology and innovation — although tensions between the technology and trade ministers and Trump officials still bubbled up.

White House OSTP Director Michael Kratsios said Anthropic had volunteered access to its new model Fable 5.1 for testing and evaluation, while Anthropic co-founder Tom Brown, during his fireside chat with Commerce Secretary Howard Lutnick, said he “really loved” the president’s pro-data center comments on Truth Social.

The chat was part of the G20 “innovation” gathering’s second day. Big Tech leaders were invited as part of the Trump administration’s efforts to promote US AI models, encourage as much data center construction as possible, and push back against AI regulations.

The CEOs stayed on message. Jensen Huang told the crowd of ministers on Wednesday not to regulate “hypothetical theoretical harm,” but instead to focus on “actual and pragmatic harm.” Palantir’s Alex Karp said AI certainly carries huge dangers but that he doesn’t believe in the doom message, making an apparent jab at Anthropic for exaggerating risks. Elon Musk called in Tuesday to emphasize the importance of keeping the baseline of AI commerce “default legal, not default illegal,” though it wasn’t clear what he thought was illegal that shouldn’t be.

Safety testing for models came up mostly in passing. Google DeepMind’s Demis Hassabis said the opportunity had to be pursued “in a responsible way,” stopping short of detailing the industry-funded testing body he has pushed for all year. Sam Altman called the administration’s voluntary pre-release review of OpenAI’s forthcoming Astra model a “productive process.”

There was little patience for the public pushback around data centers. Echoing the now-standard counter-argument, Lutnick called the opposition movement “propaganda by our adversaries” while Treasury Secretary Scott Bessent said that AI companies had done a “horrendous job” of explaining the benefits of their technology to the American people.

The consensus statement that everyone signed at the end, alas, contained the usual platitudes about opportunity and prosperity but none of the meaty measures needed for real international cooperation on AI risks and opportunities.

EA Risk

Doomers Might Be Right but They Can’t Be the Referees

The widespread alarm over the OpenAI/Hugging Face hacking incident certainly seemed well-justified by the after-action report from METR and Redwood Research, the two outside groups that OpenAI invited in to investigate.

We don’t have reason to doubt their conclusions or downplay the risks. But we do worry about where they sit in the debate. In addition, the New York Times reported that their access was limited and they weren’t able to review all aspects of the incident.

Three auditors went to OpenAI’s premises: METR’s Ajeya Cotra and Hjalmar Wijk, and Redwood Research’s Chief Scientist Ryan Greenblatt. Two of the three have prior experience in organizations that are closely aligned with effective altruism (EA) and so-called doomer worries about AI risks.

Cotra spent nearly a decade working at top EA non-profit Coefficient Giving (formerly Open Philanthropy), including leading its technical AI safety program and directing $25 million in philanthropic grants towards LLM evaluations. Wijk completed research fellowships at Eliezer Yudkowsky’s MIRI, Nick Bostrom’s Future of Humanity Institute, and the Centre on Long-Term Risk, which until 2023 was a project of the Effective Altruism Foundation.

Two days after the METR audit was published, Cotra shared her personal surprise at the attack’s severity on Substack and said it felt “more than 50% of the way to full-blown AI takeover”; she noted that her opinions aren’t reflective of her employer or the other investigators.

Three members of the Redwood Research team, although not Greenblatt, “paired” with podcaster Dwarkesh Patel for parts of the writing and research that went into his widely read explanation of the event, “The Rise and Fall of Agent Civilizations.”

Dwarkesh’s podcast has been described by the New Yorker as “to the doomer crowd what “The Joe Rogan Experience” is to jujitsu bros.” Both Cotra and Greenblatt have appeared on it in recent months. Dwarkesh’s piece was both praised and panned for using anthropomorphic language to describe what happened, with a schism forming on X between those who view the incident as a grave development and those who see a routine cybersecurity lapse.

Clearly the public needs reliably independent assessments of AI’s risks and benefits. Perhaps this is where Demis Hassabis’ proposed AI auditing organization modeled on FINRA could prove valuable. That said, when the idea of a “rapid institutionalization of auditing and technical assessment of frontier AI labs” was floated by OpenAI’s Head of Strategic Futures Dean Ball on Thursday, it was met with cries of regulatory capture. It’s hard for the public to know who to trust.

One Big Chart

Just 1% of AI Customers Account for Most Enterprise Revenue at OpenAI and Anthropic

That spike in enterprise demand for AI tools that we keep hearing about? A lot of it comes from a surprisingly small set of power users.

Data from the Ramp Economics Lab shows that just 1% of OpenAI and Anthropic customers on Ramp’s platform account for 80% of their enterprise revenues, based on Ramp corporate card spending. The companies in that 1% skew towards tech and AI products, unsurprisingly.

If Ramp’s data is on target, that level of concentration could be a risk for the big labs’ growth projections.

Read more

Anand Lunia, the tall, lanky, and bespectacled co-founder of the VC investor India Quotient, made a name for himself with his firm’s 2015 seed investment in ShareChat, an Indian local-language social network.

The company was the poster child for Lunia’s India-first philosophy. The 53-year-old investor had long criticized foreign companies’ alleged hegemony in India’s tech sector, calling for protectionism and localization to boost the country’s startups. He argued that companies like Google should pay more taxes and be required to list their sprawling local units.

For a time, Lunia’s locals-first approach seemed to be a winner. When India’s government banned TikTok and other Chinese apps, ShareChat and its sister apps became national sensations among India’s blue collar workers. During the pandemic, it became an investor darling too: its valuation jumped from $600 million to $5 billion in two years, drawing Google and the erstwhile Twitter as investors.

By March 2022, near the peak of that hype cycle, India Quotient’s million-dollar investment out of its tiny $20 million second fund had grown by 77x.

After that spectacular launch on the back of ShareChat, India Quotient went on to raise five funds totaling $320 million. Despite being small by modern VC standards, it has developed a reputation of spotting contrarian themes that proved profitable.

Still, India Quotient’s fortunes have been closely tied to those of ShareChat, which hopes to make a public market debut soon. Now, documents exclusively obtained by Newcomer show how the lack of an exit for ShareChat (beyond some small secondaries) and its uncertain valuation have hobbled the performance of India Quotient’s second fund, and could ding the third as well.

It’s a story of how the much-discussed power law of venture capital — almost all the returns come from a few breakout stars — doesn’t always play out in the way one might expect.

The documents cover India Quotient’s performance from 2022 through 2025. They show how the firm reacted to whipsawing pandemic-era valuations — implementing a massive across-the-board write-down at one point — and how the ups-and-downs of ShareChat drove performance.

It was an all-too-familiar scenario during that period, and the fallout still haunts the return metrics of many a VC firm.

Read more

Venture capitalists are betting that a barrage of AI agents will soon arrive and upend the marketplaces for e-commerce and payments.

Most of the action today is in building the infrastructure and tools for the day when agents can be trusted with a customer’s credit card and shopping choices. How close we are to that day is one of several very big questions facing the nascent AI commerce industry.

Our upcoming Machine Earning AI Summit will explore agentic commerce, intelligent money, enterprise finance operations, the evolving consumer and shopping experience, and the major meta-themes including fraud, risk, compliance, and security. The summit will be held on September 29 in San Francisco.

Read on for a story and a market map examining a subset of the companies who are working on agentic shopping, smart payments, and the next-gen infrastructure rails that will power transactions.

This year saw a big surge in consumers using AI in some fashion to shop for goods. A survey by Adobe Analytics found that 41% of respondents used AI for online shopping in June — a number that includes simple searches with Google’s built-in AI or ChatGPT. Around 53% of shoppers said they trust AI as a recommender as much as they trust brand websites, per an April survey conducted by Retail Dive and the commerce analytics platform Rithum.

Mark Grace, an investor at M13 who’s focused on fintech and blockchain deals, said he has noticed consumer use of AI commerce tools accelerate significantly this summer. Early adopters can now find shopping agents that “can actually go and transact on your behalf,” he told us. “I think people have become more comfortable with outsourcing some of those tasks.”

Forerunner managing partner Kirsten Green (pictured above) said in a text exchange that it’s still too soon to see hard data points on how much agentic shopping has taken off, but that anecdotally more consumers feel they’re willing to trust AI tools.

There are several categories of startups that aim to help AI agents be smart shoppers and securely make payments. We’ve broken them down into five buckets below.

Multipurpose Personal Agents

Personal agents like Instinct and Town promise to perform a range of tasks including scheduling appointments and sending emails. Shopping is also a key component. Early users reported Instinct landing dinner reservations for popular restaurants and movie tickets to special-format screenings of The Odyssey.

These personal assistants, both of which recently closed big funding rounds, are still prone to errors that have given people pause. An Instinct customer on X wrote that his agent canceled his flight without sharing that there would be hardly any refund, costing him around $300. Issues like these will certainly have to be figured out before regular consumers trust an agent with their credit card information.

Shopping Assistants

The AI shopping assistants more directly aid consumers in discovering new products from retailers, rather than having an agent spend money for them. One way to do that is through retail partnerships: Julie Bornstein’s startup Daydream, a personalized AI shopping search engine, started embedding into its clients’ websites at the end of July, so now shoppers can use the search function directly on the website to shop for that specific retailer’s catalog. Investors in Daydream include Forerunner, Index Ventures, GV, and True Ventures.

One investor we spoke with remains bearish on AI shopping agents, despite the category picking up with early power users. Shopping is something people enjoy in their free time, and e-commerce is still just a portion of the global commerce market, the investor said, so consumers will continue to choose to individually shop offline.

There’s another cautionary tale in Phoebe Gates’ shopping assistant Phia, which uses AI to find deals on online purchases. It’s been embroiled in a cookie-stuffing scandal, where it claimed referrals to certain partners that it didn’t actually earn. Phia denies the claims and disputes Bloomberg’s reporting on the topic. The company has raised around $43 million to date, and counts Kleiner Perkins, Khosla Ventures, and Notable Capital among its backers.

Marketing Tech

New types of AI-powered marketing technology are another opportunity. Venerable search-engine optimization (SEO) approaches are now being challenged by GEO — or generative engine optimization. Profound, a buzzy startup in this space, raised $96 million in Series C funding earlier this year, led by Lightspeed Venture Partners, with Sequoia Capital, Kleiner Perkins, Evantic, Saga Ventures, and South Park Commons all participating.

Smart Payments

There’s also the business of making all of these agent-led payments run smoothly. Smart payments startups like Catena Labs are setting up the banking infrastructure for AI agents by giving them digital identification to match them with the person on the other side who wants a transaction to go through. Another startup doing similar work is Basis Theory, which helps agents make payments by creating secure short-term virtual cards for agents to use for added security.

Agents buying on behalf of enterprises has excited many VCs but hasn’t had as much traction yet: Many large businesses are still too concerned about the risk from compliance failures or security breaches, said Grace.

Infrastructure

Agentic commerce is still a rounding error today, but a lot is happening behind the scenes to develop the necessary infrastructure and technical protocols to make it all work. OpenAI and Stripe have launched a framework called Agentic Commerce Protocol (ACP). Google owns and operates the Agent Payments Protocol, which can integrate with several other systems for agents to purchase products online. Visa and Mastercard also run systems called Trusted Agent Protocol and Agent Pay, respectively, to integrate AI purchasing into traditional financial services.

The protocols themselves are somewhat unsettled: Instant Checkout, the first consumer product built on ACP, launched in September 2025 but was pulled back in March, with OpenAI saying it would let merchants use their own checkout while it focused on product discovery.

Meanwhile, Stripe is building underlying infrastructure for agents to make purchases online through stablecoins, while Ramp launched AI agents to handle routine business procurement and fraud detection.

We compiled a list of over 35 startups building AI agents for consumer shopping, smart payments for businesses and customers, GEO and agent marketing tools, and agent financial infrastructure to keep on your radar, along with their notable backers. It’s based on our own research and conversations with investors in the space and focuses on earlier-stage companies.

The Machine Earning AI Commerce Market Map

Read more

Nvidia again reported soaring revenue and earnings, and acquired Hugging Face and much of Poolside, generating two big VC exits while raising questions about the company’s growing reach. Top tech companies issue an urgent call for action on AI cyber risks as the OpenAI hack incident gets more scrutiny. Sam Parr talks strategy behind big media exits on the podcast. Decent earnings from Salesforce sent its stock up more than 20% and dented the SaaSpocalypse story. Space tech investing is white hot. America’s Generalist and China’s Dogotix raised big rounds as investors swarm to the robotics sector. Meta agreed to pay up to $17.1 billion and make changes to its products to settle a lawsuit from state AGs — and put the screws to its competitors in the process. Y Combinator draws the FTC chair’s ire with a snarky DEI tweet.

The Main Item

With Hugging Face, Poolside & Incredible Growth, Nvidia Expands Across the AI Stack

After a remarkable few weeks of blowout earnings, multi-billion-dollar acquisitions, and potentially hundreds of billions in loan guarantees, Nvidia more than ever sits as both the technical and financial engine of the entire AI economy.

Given the sheer scale of its profits and its growth forecast, the company could well be up for the task. It earned $54 billion last quarter on revenue of $96 billion. CEO Jensen Huang projects 70% revenue growth for next year. Never has a company anywhere near that big grown anywhere near that fast. (By comparison, Google’s growth rate only once reached as high as 40% after it crossed the $20 billion annual revenue mark in 2008).

And yet. The scale of Nvidia’s ambitions is also off the charts — and raises tricky questions for customers, investors, and policy-makers alike. Its $6 billion acquihire of much of the Poolside team, which we scooped last week, is aimed at building open-source models that will inevitably compete with its biggest customers. Thursday’s report that it would acquire Hugging Face for $13 billion furthers its grip on tools that are used by all AI companies. Nvidia’s reported interest in investing in Perplexity at a $30 billion valuation confirms its willingness to underwrite a wide swath of the industry.

On top of that, the loan guarantees that it’s extending to customers including OpenAI, potentially totaling hundreds of billions, are raising alarms about circular financing. Still, oft-cited precedents such as the fiber-optic bubble of the early 2000s don’t necessarily apply.

Clearly, the markets for now are taking Nvidia’s results as a signal that the AI boom has plenty of room to run. Certainly Hugging Face investors, led by Lux Capital and including basketball superstar Kevin Durant, were celebrating. “Investors are sharing ~$7.3bn of profits on <$400m invested. A cool ~20x blended,” noted Paul Bonnet in a tweet breaking down the winnings.

But of course there were Nvidia skeptics too. One notable theme among the comments: Nvidia is leaning into open source, and thus setting up to compete with its customers. Maybe that’s because OpenAI, Google, Amazon, and even Meta are investing heavily in their own chip-development efforts that will eventually compete with Nvidia.

Here’s what people are saying:

• “The “AI capex is a bubble” and “SaaS is dead” narratives getting shredded this morning. NVDA +8%, SFDC +20%.” — former White House AI advisor David Sacks.

• “Nvidia just paid $12.9 billion for a company famous for giving away AI models for free….The price looks insane until you see who Nvidia is defending against.

  Its biggest customers are all building escape routes. OpenAI is designing chips with Broadcom. Anthropic trains on Amazon’s Trainium. Google has run its own TPUs for a decade. The labs writing Nvidia the largest checks are the ones working hardest to stop.” — Aakash Gupta

• “Every day i read about some chip that is superior to Nvidia. And every year i see no real competitors” — Jim Cramer

• “Here is another example of the vendor-financed demand story: 1. Nvidia holds equity in CoreWeave 2. THEN sells CoreWeave the chips 3. AND backstops them with $6.3 billion valuation 4. WHICH CoreWeave takes to a bank (signed customer contract) 5. TO BORROW AGAINST and buy more GPUs. Yes, it’s legal but that doesn’t make it any less round tripping!!” — Samantha LaDuc

• “Open Source is winning bigly. Tokens will be 100x cheaper in 24 months. You’re going to run 50% of tokens on your local hardware unmetered. @Dell, @nvidia and @Apple are the big winners in this trend” — Jason Calacanis

• “$NVDA, like the US government, keeps making promises it cannot keep while going deeper into debt. Absolutely insane that people are focused on revenues instead of looking a few paras below the headline.” — Kashyap Sriram

• “Kevin Durant and his business partner, Rich Kleiman, invested in Hugging Face’s seed & Series A fundraising rounds. Nvidia just acquired the company for $12.9 billion, and since I’m told Durant invested $100,000 in the seed round and $150,000 in the Series A, he likely made more than $60 million on this investment alone. That is one of the best athlete investments ever (and more money than Durant will make playing in the NBA this season).” — Joe Pompliano

The Big Cheese

We told you we’d tell you when it happened: The Big Cheese is coming to San Francisco on Sunday, Sept 13. It will be screening at the Roxie. Buy tickets here.

Red Alert

Industry Leaders Issue Urgent Call for Collective Action on AI Cyber Risks

The OpenAI-Hugging Face agent-hacking incident, which ongoing analysis shows to be even more frightening than initially believed, has apparently served as a wake-up call on AI risks, at least as it pertains to cybersecurity. An unprecedented coalition of leading companies on Thursday issued a joint statement saying there was a short window of a few months to harden infrastructure against a blitz of AI-powered cyber attacks.

The statement was, on the one hand, lacking in specifics. On the other hand, it evangelized for much-needed, common-sense collaboration among industry players and the government to defend against the expected agent-hacking wave.

Also this week, Bill Gates took an even more ambitious swing on AI safety, calling for a systematic all-of-society effort to prepare for the massive disruptions that AI is likely to bring.

Anyone doubting the need for any of this should read the after-action reports on OpenAI-Hugging Face hack. An agent swarm engaged in exactly the kind of deceptive, scheming behaviors that have long been feared, ignoring clear human instructions in favor of pursuing its goals at any cost. And it was only discovered by accident. OpenAI released its investigation Wednesday.

Ezra Klein goes deep on it in this podcast with Helen Toner. It isn’t pretty.

SaaS Comeback

Salesforce Rebounds as Anthropic Bet Shines

There are few things Salesforce loves bringing up more than the fact that it owns 1% of Anthropic. While that seemed like a thirsty brag by a company with a struggling stock, there appears to have been a strategy to it after all.

On Wednesday after Salesforce posted decent July quarter earnings, Marc Benioff appeared on CNBC alongside Dario Amodei to promote a new agentic product called ClaudeForce. The company’s stock shot up 22% the next day.

It was a remarkable jump considering the company’s numbers were merely decent — 11% growth from the prior year and a slight beat on expectations.

The bigger story here was that the long-predicted SaaSpocolypse has yet to pass. There was no major die-off in Salesforce’s customer renewal rates. The company isn’t killing it, but it’s not being killed either. And if Claude is positioning itself as a partner rather than a replacement, then SaaS lives to fight another day. With the jump in share price, Salesforce is now flat for the year. We’re back to where we started.

Meanwhile, Nvidia’s deal spree is building toward the company’s so-called Nemotron project of creating a collection of open source agentic AI models, which would challenge the hegemony of the frontier labs. This again plays into the hands of the CRM companies like Salesforce, which can promise customers their choice of models rather than locking them into a single provider like Anthropic.

That the pendulum swings on all these companies on a weekly basis speaks to the fact that no one really knows how this will shake out. But this week showed that the loudest SaaS doomsayers have been quieted for now.

Newcomer Podcast

Sam Parr on The Hustle, Hampton & Why the OpenAI-TBPN Deal Was a Mistake

One Big Chart

Space Tech Deals Top Last Year’s Total

Startups building for space have been white hot this year. Fresh off SpaceX’s record IPO, global VC funding for space tech deals for the first two quarters of 2026 has already surpassed the total deal value for 2025, with $11.3 billion across 244 deals, a new report in PitchBook shows.

Last year, space tech startups raised $10.1 billion invested across 433 deals.

Read more

Update on Sept 3:

Go read our friend Alex Konrad’s story for a major new twist in the personal agent saga.

A source confirms to us that Forerunner and Menlo are replacing Index as the leads of the Town round after potential rival Instinct appeared upset that Index planned to invest in both companies.

Konrad writes, “To be clear, our friends at Newcomer didn’t have the initial story wrong, our own sources say: this was an about-face that is unusual for such a buzzy round and major investors involved.”

Below is our original story.

Two startup personal assistants have venture capitalists in a tizzy right now:

• Town, the affable enterprise personal assistant, is in talks to raise at a $1 billion valuation in a round led by Index, sources tell us. The corporate personal assistant, started by former Plaid CTO Jean-Denis Greze and former Google applied AI product director Tony Vincent, received earlier backing from Andreessen Horowitz and Forerunner.

• Meanwhile, Instinct, the text-based consumer personal assistant run by the company Spear Street Technology, is raising round after round. Founded by former Sierra researcher Noah Shinn, it has cumulatively raised $350 million with investors valuing it at $2.5 billion, as the Wall Street Journal first reported Wednesday. Index Ventures and Benchmark are leading this most recent round.

They’re two of the hottest startups around with venture capitalists eager to invest more money.

We were surprised that Index is backing both startups at the heart of the personal agent frenzy — though as one investor pointed out to us — Greylock once invested in both Facebook and LinkedIn.

Right now, venture capitalists themselves seem like the power users of these products. But investors see the possibility for generation-defining applications.

Read more

It was mostly by chance that the acquisitions of Cursor and OpenRouter landed just four days apart. But for a16z’s infrastructure team, which had a hand in both, it made for an especially glorious victory lap: Cursor was the biggest M&A exit of all time, and OpenRouter was a monster too, returning more than 50x in just two years.

So when Martin Casado, who leads that team, reached out to talk shop after the news, we were surprised that what animated him most was all the maneuvering to get credit for the two big scores.

Individual VCs claiming deals, he told us, is “a historical relic” of a bygone era in tech investing. Modern venture, he argued, doesn’t rest on individual decisions.

“I’ve done almost 200 deals,” Casado said. “I can’t think of a single deal that other people weren’t absolutely instrumental in, whether it’s closing the deal, sourcing, doing the picking, diligence.”

Read more

Between Cursor and OpenRouter, a16z’s infrastructure practice had a week for the venture history books. OpenAI and Anthropic show strong revenue figures, but hang-ups remain ahead of their planned mega IPOs. Sequoia tops the leaderboard for most investments in newly minted unicorns. A missile startup raises a 10-figure funding round while chip deals stay white hot. A new report says Nvidia is ramping up special chips for China. American public sentiment on data centers hits its lowest point yet.

The Main Item

Everyone Wants to Claim Credit for Cursor & OpenRouter. They’re Among the Biggest Scores in Venture History.

To power a megafund, you need mega-exits. Thankfully for a16z, they are suddenly flowing in, big time.

• SpaceX this week closed its acquisition of Cursor for $60 billion in stock, the largest buyout ever of a venture-backed startup.

• Stripe agreed to acquire OpenRouter for around $8 billion, per our sources, in what Axios reported to be a mix of cash and stock.

The value of a16z’s combined outcomes in the two companies is north of $8 billion, on total investments that our sources pegged at about $320 million. The speed at which they happened was also extraordinary.

Behind the victories sits the firm’s infrastructure practice helmed by general partner Martin Casado, who oversaw term sheets for both Cursor and OpenRouter and held a seat on Cursor’s board at the time of the acquisition.

Both deals were buzzy enough that people were jockeying for credit on X all week, with a16z partners comparing Casado’s hot streak to Michael Jordan in the 1990s. A16z first backed Cursor as the lead on its 2024 Series A, then co-led its Series B with Thrive shortly after.

Many investors gave kudos to a16z infra GP Matt Bornstein for initially leading the Cursor deal. Casado acknowledged his role in a post on Bornstein’s promotion from late January.

Subscribe now

Anjney Midha, formerly of the a16z infra team, scored the lead slot on OpenRouter’s seed round in 2025, with Casado on the investment memo, as we detailed Thursday. But Midha left the firm shortly after that deal and a16z has publicly stressed the role of Casado and Chris Dixon in making the OpenRouter investment happen.

A16z this week also had some good news in the form of the crypto market bounce, with Bitcoin back up over $70,000 for the first time in many months. OpenRouter, in fact, traces its origins to the crypto business — co-founder Alex Atallah was building a marketplace for digital assets before pivoting to a marketplace for LLMs. And acquirer Stripe is keenly interested in tokenization and has already bought various crypto companies.

Too Big To Score?

There’s been a debate about whether megafunds could truly deliver VC-level returns at such hefty scale. A16z has always been at the center of that debate: its initial $300 million fundraise in 2009 was large enough to reset expectations for what a venture fund could be, and its funds have only continued to balloon.

Yet per Newcomer’s reporting last fall, that first gamble paid off spectacularly well: it returned 6x in DPI and sits in the top 5% of funds from that year.

These days, a16z raises more than quadruple that amount of cash for just a single sector. The firm brought in $1.25 billion for its first dedicated AI infra fund in 2024, as part of a $7.1 billion overall fundraise. It quickly followed that with $1.7 billion more at the beginning of this year, as part of its $15 billion capital haul.

Now a registered investment advisor with a variety of late-stage vehicles alongside more traditional venture funds, a16z is a diversified asset manager with scale that’s starting to rival big Wall Street firms. (In March, a16z filed that it had $106 billion in assets under management.)

Still, most of its post-2016 funds had failed to deliver meaningful DPI to limited partners as of the end of 2024, according to the return data obtained by Newcomer. This week’s deals, on top of 2025 wins with Figma and Wiz, would certainly change that.

At the same time, the firm is also a poster-child for the idea that exits aren’t everything anymore given the growth of private markets. So far, LPs have been willing to go along with that, taking paper gains on Stripe and SpaceX and many others. Now Stripe is showing it can use its shares as currency public-company-style, which works partly because the company strives to make secondaries possible.

SpaceX of course is public now, though it’s unclear if and when various investors can unload their shares, even if they want to.

Too Big For the DOJ?

A16z’s reach across the infrastructure stack has drawn some less welcome attention. Bloomberg reported the Justice Department is investigating the firm over board seats at two competing companies: Ben Horowitz at Databricks and Casado at Fivetran.

These probes don’t always produce action, but a case against a16z could make it harder for megafirms to spread their bets across the AI stack.

Still, for the moment at least, that’s a small worry next to the big wins. Like Thrive, which we wrote about last week, a16z is showing that innovative approaches and megafund muscle are a good match for the AI era.

Subscribe now

Great Expectations

OpenAI & Anthropic’s Latest Revenue Figures Are Incredible But Worries Remain

Fresh financial details on the frontier labs this week painted a mixed picture, given the bigger context at least.

On Monday, Anthropic told investors it had annualized revenue of $65 billion, according to Bloomberg, and that it was projecting annual revenue of $190-200 billion by 2028. Then on Tuesday the Wall Street Journal reported that OpenAI had recently told investors it generated $6.7 billion last quarter — up 18% from the previous quarter and significantly below the $11.5 billion that Anthropic reportedly did during the period.

On their own, these numbers are astounding for companies that are barely three years into earning revenue. But considering the pumped-up IPOs and valuations they’re gunning for in the coming months, they also raised some red flags.

The biggest were for OpenAI, which scrambled to counter the narrative that its growth was tepid. The day after the Journal’s story, OpenAI CFO Sarah Friar told employees that the company’s revenue run rate is up 35% quarter to date and that its enterprise revenue run rate is up 50% quarter to date, according to CNBC.

Even Anthropic’s amazing annualized revenue fell a bit short of some of the craziest expectations, like YipitData’s projection of around $80 billion annualized revenue. All these figures will be scrutinized intently as they march toward IPOs in the coming months, carrying much of the AI industry with them.

Anthropic may file publicly as soon as this month for an offering that’s likely to smash the recent SpaceX record for biggest-ever IPO, Bloomberg reported. As part of the deal, Anthropic has been working to give Dario Amodei and other co-founders a new class of stock with extra voting power in order to shield them from external shareholder pressure, according to The Information. A complicated organizational structure can have its… complications.

Machine Earning

Apply to Secure Your Spot at Our Machine Earning AI Summit on Sept. 29 in San Francisco

We’re five weeks out from Newcomer’s first-ever Machine Earning AI Summit in San Francisco on September 29.

We’re bringing together the most prominent venture-backed founders and builders in commerce and AI for a curated invite-only event.

Spots are extremely limited, so apply as soon as you can for a chance to attend.

Apply to Attend

One Big Chart

Sequoia Has Backed the Most Unicorn Startups So Far in 2026

It’s been quite a year for startups winning unicorn status: some 250 have reached a $1 billion valuation or higher, up from 193 for all of 2025.

Per fresh data from Crunchbase’s Unicorn Board, Sequoia tops the leaderboard for most portfolio companies in this new crop of unicorns, with 52 investments in 29 companies.

Read more

EXCLUSIVE TO NEWCOMER: Poolside AI, the artificial intelligence model-building startup, has struck a non-exclusive licensing deal with Nvidia for $6 billion, plus a $1 billion investment in Poolside at a $12 billion pre-money valuation, according to a letter to investors obtained by Newcomer.

Read more

OpenRouter’s founder Alex Atallah and AI investor Anjney Midha became fast friends over 13 years ago at Stanford, but it took more than a decade for them to partner up professionally — and the results have proven spectacular.

Stripe announced Wednesday its agreement to purchase the model routing platform, co-founded by Atallah in 2023. Midha had backed the team in 2025 while he was a general partner at Andreessen Horowitz. It’s the type of rapid-fire founding-to-multibillion-dollar-exit story the AI boom has supercharged in Silicon Valley.

We got the details on the deal and OpenRouter’s cap table from sources close to multiple firms with stakes in the company, all of whom requested anonymity to discuss sensitive financial information. A few key numbers:

Read more

Last week, in New York City, I dropped by the offices of AI-native law firm Crosby to listen in on a small summit organized by investor Jake Saper, who is spreading the gospel of AINS, or “AI-Native Services.”

Saper and his firm Emergence Capital are betting that accounting, insurance, legal, and other people-intensive industries will be the next frontier for startups.

Instead of selling software like last-generation startups did, these next-generation AI startups are selling the underlying services themselves. They’re hiring lawyers, accountants, and insurance brokers and pairing them with AI-pilled software engineers who are tasked with figuring out ways to make those service providers much more productive. These AINS businesses are often charging for the results of their services rather than billable hours (because they aim to do the job in far fewer hours than traditional human-powered services businesses could). So you might pay per completed contract, say, rather than for the hours a lawyer says it took.

I caught up with Saper after the summit, and he sent me the private deck he shared with eager AINS founders that contains his emerging playbook for building these businesses.

You can see the key slides for the AI-Native Services playbook at the end of this post.

Read more

Thrive Capital raises a fresh $2 billion for Thrive Holdings and will own a piece of the Los Angeles Lakers. Prediction markets are on a roll as Kalshi and Polymarket seek fresh funds at $40 billion and $20 billion, respectively, but we’re not fans of the category. Open-source AI is still scarce in the enterprise, despite all the hype. XAI co-founder Igor Babuschkin raises $1.1 billion for his new venture. European startup stars Lovable and Cambridge Aerospace close big Series C rounds. Brad Lightcap becomes the latest top exec to leave OpenAI. Phoebe Gates faces fresh allegations of startup deception. The White House targets OpenAI policy advisor Dean Ball. Accel raises $3.5 billion for early-stage deals. Former General Atlantic investor Anton Levy brings in $1.1 billion for Layer Global.

The Main Item

Hard Work, OpenAI & Trump Ties Paying Off for Josh Kushner

Thrive Capital’s big moves this week exemplify how founder Joshua Kushner has steered the firm far afield of typical Silicon Valley venture capital norms. It’s been working out well so far.

Thrive is headquartered in the historic Puck Building in lower Manhattan, far from Sand Hill Road or downtown San Francisco. It has raised funds for private equity-style roll-ups and invested in an indie film studio. Kushner’s status as the brother of President Trump’s son-in-law Jared has given him proximity to an extraordinary center of power beyond the typical business sphere.

Wednesday’s announcements were both striking examples of Kushner’s increasingly expansive ambitions:

• Thrive’s AI roll-up vehicle Thrive Holdings closed a $2 billion outside fundraise at a $12 billion valuation. Investors in the funding round included SoftBank, Altimeter Capital, and D1 Capital Partners.

• Founding partner Josh Kushner told press in a joint statement with Bob Iger the two agreed to buy out Mark Walter’s 80% stake in the Los Angeles Lakers, partly through another vehicle, Thrive Eternal.

For years, Thrive was a bit hard to define: it was round-agnostic and trying to muscle its way in on more established Silicon Valley firms. Now it has three clear core theses.

The first is closest to traditional venture, with a focus on AI and infrastructure via the main Thrive X fund, with $9 billion allocated for growth and $1 billion for early stage.

The second is a bet on AI transformation inside companies that it hopes to accelerate through Thrive Holdings. Its flagship investment in OpenAI feeds into this: OpenAI owns a stake in Thrive Holdings, and OpenAI employees can take on roles that sound awfully similar to forward-deployed-engineers at the companies Thrive aims roll up and empower with AI tools.

The third is a play that live events and cultural institutions will become a scarce and increasingly valuable commodity, the thesis behind Thrive Eternal. While the fund stumbled in its efforts to buy into the World Cup after a ferocious global backlash, a piece of the Lakers will be a good salve for that wound. (NBA rules limit investment funds to 20% ownership and Kushner and Iger will make large personal investments.)

Thrive Eternal already has a small stake in the San Francisco Giants, and Thrive Capital’s main fund invested in indie film studio A24 back in 2024.

In an investor letter Thrive sent out earlier this week, and obtained by Bloomberg, Kushner essentially laid out the multi-pronged thesis. He nodded at plans to invest more often in public companies, explaining that “the more precise our understanding of public markets, the better our judgment will be in private markets.”

The letter also disclosed the returns of the main funds, which included a 2022 vintage marked at 10.5 times gross — one of the better numbers in venture this decade. The letter also quietly showed that the last Thrive fund to return meaningful cash to its investors was raised in 2016. Everything since is paper markups as they wait on returns from OpenAI, Stripe, and Anduril, and on a SpaceX position that has fallen substantially since the letter’s June 30 cutoff and remains partly locked up.

Sources we spoke with give Kushner a lot of credit for the firm’s success, citing a relentless work ethic and drive to succeed. When Thrive’s main bet was on Oscar Health Insurance he was regularly grinding it out attending health conferences. As the fund has grown, he’s surrounded himself with an aggressive group of young partners and encouraged spirited debate about strategy.

Newcomer reported back in February Thrive had returned $4.3 billion back to LPs as of June 30, 2025. If the OpenAI IPO goes through even on a delayed schedule next year, Kushner will likely generate billions for his LPs.

Those results have bought Thrive the ability to keep growing. Yet rather than going all-in as a venture megafund, Kushner is using the firm’s relationships and access to capital to push into businesses that haven’t traditionally been a target for VC investors.

In that sense, Thrive may be less a venture firm that has wandered outside its lane than a firm trying to erase the lane altogether.

Cerebral Valley Show

The Most Bullish Man in AI Just Blew Up His Fund

Ethics Alert

Prediction Markets Are a Bad Bet

Silicon Valley’s era-defining technologies often sit on a razor’s edge between utopian and dystopian.

The sharing economy promised communal exchange at the tap of a smart phone — or risking it all by jumping in a stranger’s car, or trusting others to deliver. (It largely ended up being more banal: convenient, smartphone-enhanced regulatory arbitrage.)

AI is giving everyone a genius at their fingertips — or creating a super-intelligence that will put us out of jobs and maybe destroy the world.

Juul was a way for cigarette addicts to get off the hard stuff and do something safer — until it became clear it was instead another gateway to addiction, creating a generation of young vape fiends.

Similarly, prediction markets seemed to promise opportunity and insight, a new way to see into the future thanks to the genius of the crowd. But quickly the reality has set in: Kalshi and Polymarket, which dominate the industry, are primarily places to gamble on sports. Some 80 percent of the trading volume on market leader Kalshi is sports betting.

On top of that, a host of deeply problematic issues have arisen around wagering on events where motivated people can potentially game the results.

As these companies raise more and more money — just this week, Kalshi was reported to be raising at a $40 billion valuation and Polymarket was said to be doing the same at $20 billion — Silicon Valley has to ask itself: if this is really the world that it wants to build? At a moment when there are so many genuinely revolutionary AI companies getting marked up at lightning speed, why spend energy getting young men to waste more and more of their time and money on gambling?

There is a world in which Kalshi and Polymarket and the fleet of emerging rivals serve as truth machines for discerning the wisdom of crowds, and markets for hedging various sorts of risk.

That is not our world. The reality is that the main business of prediction markets is deregulated sports betting, essentially a workaround to state gaming laws that mostly bar wagering by anyone under 21, and ban it entirely in 11 states.

This is legal thanks to the efforts of Michael Selig, the crypto and prediction industry lawyer whom President Trump named to head the Commodity Futures Trading Commission (CFTC). Selig insists, implausibly but with the power of the government behind him, that the ability of American businesses to freely hedge their risks requires that Washington override all state regulation of sports betting. Selig this week invoked the CFTC’s emergency powers to block legal efforts by the state of New York to ban Kalshi for violating the state’s gambling laws, the latest skirmish in a nationwide fight.

It’s hard to ignore the fact that the industry enjoying such sweetheart regulatory treatment features a highly unusual arrangement where Kalshi and Polymarket, otherwise fierce rivals, both count Donald Trump Jr. as a close advisor (and in the case of Polymarket, the president’s son is also an investor through his 1789 Capital fund).

People have already taken hairdryers to thermometers to win prediction market wagers, which is funny-not-funny in the context of betting on military maneuvers, or wildfires, or flight cancellations. Insider trading of various sorts appears to be rampant on the platforms, despite belated clean-up efforts.

As with crypto, prediction markets are fun sandboxes for experimenting with financial structures and trading strategies and theoretically could yield some social goods. Theoretically is the operative word here though.

It’s hard to see how the current anything-goes national regulatory regime could survive a Democratic administration, or even a non-Trump Republican one. Regardless, let’s not wait for the horror stories about broke young families or people dying for someone’s parlay. There are plenty of reasons to think twice before joining the prediction market party.

Machine Earning

Apply for a Spot at Our Machine Earning AI Summit on Sept. 29 in San Francisco

Newcomer is hosting our first-ever Machine Earning AI Summit in San Francisco on September 29.

We’re bringing together the most prominent venture-backed founders and builders in commerce and AI for a curated invite-only event.

Spots are extremely limited, so apply as soon as you can for a chance to attend.

Apply to Attend

One Big Chart

Open-Source AI Makes Gains But Still Lags Far Behind in the Enterprise

Despite all of the chatter about open-source models in recent weeks, they still account for a very small share of enterprise AI use, though they’re growing steadily from their tiny base.

Read more

When Starcloud, a space data center startup, announced in March that it had raised $170 million at a $1.1 billion valuation led by Benchmark and EQT, you had to read deeply into the press release to learn that the deal was done in two tranches, with the first part led solely by Benchmark.

What it didn’t say: that first piece came at a valuation of $250 million, according to two sources familiar with the deal. Just days later, the second tranche, which also included a host of smaller investors, closed at more than four times the price.

The Starcloud deal is an example of what investors say is becoming a common practice: prestige firms getting a significantly better price than other investors in what is essentially the same round.

Proponents of such deals say they’re simply reflecting the market reality that some investor dollars are greener than others and it’s only logical for firms to cash in on that — especially in early-stage deals where an investor’s brand name can make a huge difference. But critics say such structures can be problematic, not least for employees.

Brendan Foody, CEO of the AI-training startup Mercor, stirred the pot on the issue on X last month, writing: “in the last 6 mo’s ive seen a half dozen rounds where sequoia invests in 2 tranches. everyone pretends they only did the higher valuation,” and calling the tactic “deceptive.” Sequoia partner Shaun Maguire, in a reply, claimed the practice was rare. But Newcomer’s reporting suggests it’s increasingly common across the industry.

Weston Moyer, managing partner at MVP Ventures, an investor in defense tech firm Anduril and voice AI startup Wispr Flow, says about 25% of all deals he’s seen in recent months have featured dual valuations for essentially a single funding round.

That estimate was echoed in interviews with six other early-stage VCs, who said the mechanism has gone from rare to pervasive in the current frenzied funding cycle.

Read more

We’re less than two months out from the inaugural Machine Earning AI Summit, and spots are filling up fast.

If you’re a top founder or investor focused on the intersection of AI and commerce, you can apply on our site here.

Apply to Attend

Today, we’re announcing three new speakers to our summit lineup.

Our newest speakers are:

• Kirsten Green, founder & managing partner of Forerunner Ventures

• Katie Haun, founder & CEO of Haun Ventures

• Ilan Zerbib, founder & CEO of Sapiom

  

All three bring firsthand knowledge of how AI is reshaping the foundations of commerce and finance. We can’t wait to have them speak.

They’ll be joining our stacked lineup, which already includes:

• Omer Ismail, founder & CEO of OnePay

• Jackie Reses, founder & CEO of Lead Bank

• Jack Zhang, founder & CEO of Airwallex

• Max Rhodes, founder & CEO of Faire

The Machine Earning AI Summit is a one-day, invite-only event covering the ins and outs of the new AI economy. It will be attended by top founders, investors, media, and other leaders in AI and commerce.

We’re going to be diving deep into topics like how AI is changing everything from banking to the underlying rails for payments, and what online shopping looks like when agents are the ones doing the spending.

Tickets are $699. Founder tickets are $99. Professional Services and Business Development tickets are $999.

In addition to our flagship sponsors Airwallex and Lead Bank, we’d like to thank our newest sponsors Exa, Nebius, and OnePay.

We’ll be announcing more speakers and discussion group leaders in the coming weeks.

If you’re interested in sponsoring or talking to us about getting involved, email summits@newcomer.co.

We can’t wait to see you there.

Apply to Attend

Jesse Zhang says the case for AI customer service is obvious. That’s what makes it an easy sell. It’s also what makes it a crowded one.

Decagon has crossed $100 million in annualized revenue, the company told me — in AI customer service that’s far from first place. Zhang’s plan to vault ahead of competitors like Sierra and Salesforce is speed. The Decagon CEO wants his customers to move fast: the product should be intuitive, quick to customize, and require minimal handholding.

“We are in a use case where speed matters a lot,” Zhang told me. “The way we’re going to prove it to you is that you’re able to go live way faster and you’re able to make changes and hill climb the quality of the agent a lot faster.”

(That theory of business resonates with a guy who graduated from Harvard in three years and got married at 24.)

As part of that pitch, he and his three-year-old company have taken a strong stance against being wedded to forward deployed engineers (FDEs), whom vendors embed inside customers to customize and manage software for them. Despite the rising popularity of the role across tech, Zhang sees reliance on it as a sign of a weak product — and as a place where Decagon stands out from its bigger and better-funded peers.

Sierra is an 800-pound gorilla that has three times the funding, a Valley-famous CEO in Bret Taylor, and has well-surpassed $200 million in annualized revenue, according to a source close to the company. Salesforce is frantically pivoting into AI and boasting success: it recently disclosed that Agentforce surpassed $1 billion in ARR. In June it doubled down, agreeing to buy Fin, the customer agent company formerly known as Intercom, for $3.6 billion.

Decagon and its investors — which include Bain Capital Ventures, Accel, and a16z — argue that the product, not a large services organization, is what landed major airlines and credit card companies as customers.

Read more

In 2020 my wife, Sara Joe Wolansky, helped to convince me to give up my job as a reporter at Bloomberg to start this newsletter.

Two years later, I had her back when she decided to quit her full-time job making documentary shorts for the New Yorker to pursue a yet-to-be-identified first feature documentary.

NYC Tickets

I’m proud to share that her film, The Big Cheese, a feature-length documentary about American cheesemongers, is having its theatrical premiere starting tonight at the IFC Center in New York City.

In a wonderful review this week, the New York Times called The Big Cheese a “satisfyingly tender story,” writing, “This is ultimately a tale of resilience, and Wolansky focuses on a kind of nourishment that transcends food: the power of community.”

Other early reviews have called it “bright, tangy, and bold,” and “thrilling and intense.”

From my point of view, The Big Cheese is a gripping, energetic, and loving film that will have you laughing, crying, and hankering for Challerhocker.

I played a small role as an associate producer of the film, which mostly means that I was a supportive husband.

I’ll be at tonight’s screening at the IFC Center at 7:35 p.m. I’m giving away five tickets to the first requests that I receive. Reply to the newsletter if you’d like a ticket.

There will be a Q&A with Sara Joe and the film’s star, Adam Moskowitz, moderated by the one-and-only David Zabar of Zabar’s.

The Big Cheese is going to run all week at the IFC Center and then start screening in Los Angeles on Aug. 14.

(I know many of you will want to know when the film is in San Francisco. I’ll try to drop a line in the newsletter when that happens.)

I hosted a party last night with Joe Marchese at Human Ventures in New York City for friends of the newsletter to celebrate the film’s release. Thanks, Joe, for helping me to support Sara Joe.

Thanks to O’Shaughnessy Ventures for helping to fund the film and make this whole thing possible.

Both as an independent creator and as someone who writes about startups, I know how hard it is to build something out of nothing. Making an independent film requires true tenacity, and I’m so proud and excited that Sara Joe’s project is coming to theaters near you.

If you love cheese, the movies, and air conditioning — or just want to have a good time meeting a plucky new subculture — go check it out!

Bending Spoons agrees to acquire Airtable at a steep discount to its ZIRP-era valuation, but early-stage investors make out well. Jeff Dean leaves Google DeepMind and Demis Hassabis steps back amid an overhaul of Google’s AI teams. Fresh TVPI data from Carta shows how wide the gap is between a good and great outcome in venture. Hardware companies from nuclear reactors to automated manufacturing raise billion-dollar rounds. Two profiles on Sequoia dig into the inner workings of the firm under new stewards Alfred Lin and Pat Grady. Earnings calls spook the markets on SpaceX but give Palantir a big boost. Jasmine Sun travels to the Midwest to discover why Americans hate data centers so much. Nikita Bier steps down as X’s head of product.

The Main Item

Airtable’s Sale to Bending Spoons Wasn’t a Dream Venture Exit. Still, Timing Can Be Everything.

Airtable’s definitive agreement on Tuesday to sell to European roll-up juggernaut Bending Spoons at an enterprise value of $1.285 billion could be just a taste of the disappointing exits in store for pre-AI software-as-a-service companies.

Airtable, founded in 2013, had around $480 million in ARR and was growing more than 20% year-over-year at the time of the deal. That still wasn't enough in an era of supercharged AI growth and worries that agentic coding will kill software companies. Even counting the roughly $965 million in cash, that brings the deal's equity value to $2.25 billion, it's a steep discount to Airtable's pandemic-era $11.7 billion valuation.

Still, smart investors find a way to make money.

Venture can be a timing game. While Airtable’s exit may not play as well into the lifetime-partner narrative many VCs push, strategically planned cash-outs made some VCs quite a pretty penny even though the ultimate outcome was underwhelming.

Freestyle Capital GP Dave Samuel is one investor in Airtable who cashed out at the right time.

Samuel wrote Freestyle’s first check into Airtable’s first funding round in March of 2013 for about 10% ownership at a less-than-$10 million valuation before other investors came in, we heard. Freestyle followed on through the Series B round.

The two teams’ relationship went back years before: Samuel had met Airtable CEO Howie Liu when Liu was still a senior at Duke University, and Liu served as Freestyle’s first intern right after the fund’s launch. Liu’s first post-college job was at a Freestyle portfolio company.

As a general house rule, Freestyle sells some shares of its portfolio companies once they hit a unicorn valuation. Freestyle exited a significant amount of its position in 2021 at Airtable’s peak valuation of $11.7 billion. While the exact share count couldn’t be learned, Freestyle’s multiple on invested capital will be more than 55x, we heard (The Wall Street Journal previously reported on Freestyle’s multiple).

Early investors who held on will make decent returns.

Caffeinated Capital's Raymond Tonsing wrote his first check a month after Samuel, in April 2013, and led the following seed round. Angel investor Felix Shpilman also backed Airtable at seed. Tonsing didn't lead the $7.6 million Series A — CRV's Max Gazor did — but he invested, and co-led the Series B alongside CRV. He backed every round after that, through the Series F in December 2021. Tonsing, Gazor, and Shpilman will likely get a sizable return on shares they’ve retained, although we’ve heard that the final totals have not been confirmed by the company yet.

Airtable raised at a $1.1 billion valuation in 2018 for its Series C, with Thrive, Benchmark, and Coatue all participating. Every round after that priced above Tuesday’s deal. The company raised more than $1.4 billion over its 13-year history and hadn’t spent most of it. (At the time of the signing, the company had about $965 million sitting on its balance sheet.) Late investors including Greenoaks, WndrCo, XN, and Franklin Templeton should get their capital back.

Still, a lot of TVPI will have to be readjusted internally for funds who were holding Airtable at $11.7 billion.

Ahead of the deal, Airtable spun out its fast-growing AI agent platform Hyperagent.

Early angel investors who spoke with Newcomer said Airtable has not informed them whether their ownership stakes will transfer over proportionally to Hyperagent.

Despite his own success on this deal, Samuel sees the writing on the wall for SaaS businesses. He told the Journal that he’s “strongly suggesting” that all of his portfolio companies in this sector attempt to sell to at least get out while they can.

Machine Earning

Apply for a Spot at Our Machine Earning AI Summit on Sept. 29 in San Francisco

Newcomer is hosting our first-ever Machine Earning AI Summit in San Francisco on September 29.

We’re bringing together the most prominent venture-backed founders and builders in commerce and AI for a curated invite-only event.

Spots are extremely limited, so apply as soon as you can for a chance to attend.

Apply to Attend

DeepMind Shakeup

Google Loses More Top Scientists at a Key Moment in AI Race

Google shook up its AI leadership this week, making clear what people have been whispering about for months: it’s fallen behind.

On Wednesday Google announced its chief scientist Jeff Dean was leaving after 27 years to start his own company, and Demis Hassabis was stepping back from his role as the CEO of DeepMind to become the division's Chair and Chief Scientist at Alphabet, letting him focus more on Google’s AI biotech wing Isomorphic Labs.

DeepMind’s CTO Koray Kavukcuoglu was promoted to senior vice president and will report to Sundar Pichai, but notably he will not take a CEO title.

Dean, alongside three other Google researchers, is founding a neolab called Discovery Loop to chase recursive self-improvement. Radical Ventures and Khosla Ventures have already backed it, alongside Kleiner Perkins, Lightspeed, Doerr Capital, and Google itself.

The tech giant put a friendly spin on the news. Hassabis addressed employees in a memo saying he had to step back from day-to-day responsibilities with the promise of AGI fast-approaching.

But the rosy views couldn’t distract the stock market from the reality that Gemini’s next Pro model is months behind schedule, and many other top scientists have defected to OpenAI and Anthropic in recent weeks. Noam Shazeer took off for OpenAI in June, while AlphaFold Nobel prize winner John Jumper and other top scientists left to work on Anthropic’s growing scientific initiatives. Alphabet’s stock fell on the news.

Talking to sources at DeepMind, we hear the problems have been organizational as well as cultural. They’ve struggled to build a competitive coding model partly because developing products for enterprise customers is historically a weakness for Google. The company has traditionally been stronger on the consumer side of things, which is why Gemini has cut into ChatGPT’s market share. Unfortunately for Google the momentum (and money) is in coding — and its models haven’t measured up.

Despite the hand-wringing, it’s hard to count Google completely out of the AI race now. It’s been dogged on before about model performance and slow-moving culture but has been able to bounce back. Any well-funded lab is perpetually one good model run and post-training process away from being near the top of the charts. It’s also the only truly verticalized tech giant.

But even a cash-printing machine like Google has its limits. Q2 marked the company’s first ever cash-flow negative quarter due to its massive capex spend. The company is expected to spend $195 billion to $205 billion this year on capital expenditures.

Dean’s recursive self-improvement company will be worth watching given the in-house talent, but Google’s race to get back into the mix will remain a major storyline through the end of the year.

One Big Chart

VC Funds Need TVPI of at Least 3x to Be Considered Top Tier

Read more

Leopold Aschenbrenner’s AI-centric hedge fund sells its public holdings after a market rout, a warning flare for the increasingly leveraged AI industry. TBPN Host John Coogan talks tech media and life after OpenAI ownership on the podcast. SPVs outpace direct share listings as secondary market mechanisms. A Thinking Machines Lab co-founder steps down citing health issues, but quickly takes a job at OpenAI. The Trump administration bans Chinese robots. Google DeepMind’s AlphaFold team sees a personnel shakeup and a few notable departures. Live-shopping app Whatnot offers a sign of life for consumer dealmaking.

The Main Item

As Markets Wobble, Circular Financing & Leveraged Bets Ramp up the Risks

The rapid retreat of AI and chip stocks over the past couple of weeks has exposed the fragility of an AI funding boom that’s drawing capital from every corner of the financial system — and is increasingly reliant on debt.

Situational Awareness, the much-touted San Francisco hedge fund led by 25-year-old wunderkind Leopold Aschenbrenner, this week provided an object lesson in the risks. Less than two months ago, the former OpenAI researcher was being celebrated by the Wall Street Journal for his prescience in seeing the AI opportunity early and turning a few hundred million dollars into well over $20 billion in less than two years. (The firm’s net asset value reportedly reached $45 billion at its height.)

On Thursday, he was forced to sell the fund’s public stock portfolio to Citadel after margin calls had left it in a liquidity crisis.

The accelerationist-leaning pseudo-anonymous account @beffjezos called the incident a “huge hit for EA-adjacent rationalist former big lab alignment researcher kind” and many meme-d Citadel’s Ken Griffin into the grim reaper of speculative investing.

It’s not that Aschenbrenner’s bets were wrong, exactly: the infrastructure and chip plays at the heart of the portfolio, including names like SanDisk, SK Hynix and CoreWeave, are still trading far above where they were a year ago. He himself has always spoken of the AI build-out as a long-term opportunity.

But “conviction” in this case meant levering up the wagers with borrowed money, and thus declines on the order of 30% on core holdings created a sudden cash crisis. It’s a reminder of how quickly fortunes can turn in an AI trade where valuations — and thus expectations — remain sky high, and risk is often stacked upon risk.

A similar red flag in that regard comes in the form of the latest wave of multi-billion-dollar Nvidia investment deals. This week, it was a $5 billion commitment to Ilya Sutskever’s Safe Superintelligence, much of which will be used to buy Nvidia gear.

Nvidia is reported to be discussing a loan guarantee for OpenAI of as much as $250 billion, also tied to Nvidia hardware.

That’s on top of $90 billion in corporate VC investments over the last 16 months alone, per the Financial Times. Nvidia participated in 283 funding rounds between 2021 and 2025, with 85% of those investments in AI startups, according to Crunchbase.

The Bank of Jensen

In many cases, these investments involve Nvidia essentially providing the money for customers to buy its hardware. Indeed, CEO Jensen Huang has been so aggressive in funding the AI ecosystem that the company is functioning almost like a bank; with annual free cash flow nearing $100 billion and little debt, that isn’t obviously reckless.

Yet such vendor financing, though it has plenty of precedents, is what brought down Lucent and much of the rest of the telecom equipment industry in the early 2000s after the first dot-com bubble burst.

Nvidia’s defenders point out that many of its deals are collateralized by the hardware, so if the company fails Nvidia gets the gear back. In a bust scenario, though, those chips would be worth a fraction of their current value.

Huang has proven himself a tech CEO for the ages, and an AI visionary. He’s also been exceptionally good at turning that into money. But banking is a different business. And as Aschenbrenner just demonstrated, seeing around corners doesn’t always guarantee success.

Nvidia’s frenetic deal-making is just one of the ways in which the AI buildout has upended the traditional start-up financing playbook. As we noted last week, corporate venture investing has exploded in recent years, and accounted for almost 90% of all VC dollars that have gone into AI firms this year. That’s compared with less than 50% of VC dollars a decade ago.

New types of financial players are also making this cycle different.

From Wall Street to Jane Street

While hedge funds such as Coatue and mutual fund companies including Fidelity started getting involved in late-stage venture deals more than a decade ago, Situational Awareness was part of a fresh wave of investment firms encroaching on traditional VC territory.

The secretive Wall Street firm Jane Street, which traditionally made its money from proprietary trading algorithms, is on the cap tables of Anthropic, chipmakers Etched and MatX, and AI cloud infrastructure heavyweight Fluidstack. (It also had money with Aschenbrenner.)

Jane Street is interested in AI tech not just as investment, but for internal use: it invested $1 billion in CoreWeave and signed a $6 billion commitment to use its cloud platform for all of its machine learning needs. It’s also in the process of financing its own AI data center.

Big money managers and private credit firms including Goldman Sachs Asset Management, Blackstone, and Blue Owl Capital have also emerged as major funders of the AI buildout, extending loan packages in the tens of billions for data centers.

Part of this is simply a function of the enormous amounts of capital needed for AI data centers and chips. But for strategic investors, there are many other reasons to play the venture game.

Plumping up potential customers is one obvious benefit. An investment could also yield access to premier AI tools ahead of the competition, and help keep a funder abreast of the latest developments. There’s also the chance of a windfall: Google recorded an incredible $99 billion gain on its equity investments in Q2, mainly from gains on Anthropic and SpaceX.

“Enterprise software companies in particular often feel real urgency to invest in and sometimes acquire AI startups to stay competitive and defend their market share,” PitchBook senior research analyst Kaidi Gao told us.

Fear and greed continue to draw new corporate players into AI startup investing. Schneider Electric, for example, a 190-year old energy and industrials firm, is channeling capital from its €1 billion venture fund, SE Ventures, almost entirely into AI startups.

The $700 million Series A for Figure founder Brett Adcock’s new AI neolab Hark shows the new state of play: investors include Nvidia, AMD Ventures, ARK Invest, Brookfield, Intel Capital, Qualcomm Ventures, and Salesforce Ventures.

Newcomer Podcast

John Coogan on Selling TBPN to OpenAI, Sam Altman & the Future of Tech Media

One Big Chart

SPVs Outpace Direct Sales for Secondary Share Sales So Far this Year

Investors are increasingly turning to Special Purpose Vehicles, or SPVs, to purchase shares of Silicon Valley’s most sought-after private startups, rather than buying shares directly from employees or other investors.

A new PitchBook report showed that single-layer SPVs have surpassed direct-to-cap-table trades in secondary deal volume on the trading platform Caplight so far this year.

Read more