The SPV Fraud Wave Should Scare You More Than the Headlines Do

On August 10, 2026, the SEC sued Adit Ventures Management, founder Eric Munson, and three affiliated general partners, alleging the firm told an investor it already owned 32,000 shares of Klarna stock when it owned none, then separately bought SpaceX shares at $420 and resold them to a client fund at about $498, pocketing roughly $1,020,000 (SEC press release). I don't think Adit was an outlier. I think it was a preview.

What the SEC Actually Found at Adit Ventures

Start with the facts, because they are worse than the one-line summary you probably read. From April 2019 through December 2024, Eric Munson ran Adit Ventures Management out of New York, raising money across more than 60 funds from over 1,000 investors, all pitched as access to pre-IPO shares in companies like SpaceX, Klarna, and Flexport (InvestmentNews). The SEC's complaint, filed in the Southern District of New York as case 1:26-cv-06800, lays out three distinct failure modes, and I want to walk through all three because each one maps to a different structural weakness in how SPVs work.

First, ownership itself was fabricated. Munson allegedly secured more than $15 million from one investor by claiming a vehicle he controlled already held 32,000 Klarna shares. It held zero at the time. A side letter told the investor the entity "owns shares in Klarna." That sentence was false when it was written.

Second, the money moved somewhere other than where investors were told it would go. Defendants told investors capital was needed immediately to seize near-term pre-IPO windows, then, according to the complaint, used that capital for other purposes, including more than 50 unsecured loans between funds on terms favorable to Adit. Some of those loans reportedly sat unpaid for years.

Third, and this is the part I think gets underweighted in the coverage, Adit's general partners bought pre-IPO shares for themselves first and resold them to client funds at a markup, more than 150 times, without the written disclosure and consent that the Investment Advisers Act requires for principal transactions. The SpaceX example is the cleanest illustration: buy at $420, sell to your own client fund at $498, keep the spread. Multiply that pattern across 150-plus transactions and you get a business model, not a lapse.

Adit also allegedly pledged client fund assets as collateral for a $10 million line of credit, partly used to cover its own obligations, and operated for years without registering as an investment adviser, which the SEC says let it dodge routine examinations until 2024 (SEC litigation release). The defendants settled without admitting the allegations; a federal judge entered judgments enjoining future violations on August 11, with disgorgement and penalty amounts still to be set (Reuters, via Business Times). Munson denies wrongdoing and says he settled to avoid prolonged litigation, not because the claims have merit.

Why "One Bad Actor" Is the Wrong Read

Here is my contrarian argument, and I'll say plainly that this section is my analysis, not a reported fact: the SPV structure that made Adit's scheme possible is the same structure powering nearly every OpenAI, Anthropic, and SpaceX-adjacent deal being pitched to accredited investors right now. Three design features do the damage. The first is layering. Most retail-adjacent private-market access today doesn't run through a single SPV that directly holds shares. It runs through a feeder fund that invests in another fund, which holds an interest in a fund that may or may not directly hold the underlying stock. Every layer adds a place where the story "we own this" can become detached from the fact of ownership, and every layer adds a general partner who can charge a fee for being in the chain. When investors ask what they actually hold, the honest answer at each hop is often "an interest in an interest in an interest." That is not a metaphor. It is the literal legal structure of many pre-IPO access products, and it is exactly the ambiguity that let Adit tell one investor it owned Klarna shares it didn't have.

The second is the absence of an independent custodian. When you buy a public stock, a broker-dealer confirms the trade and a custodian holds the security, both regulated, both with obligations to you that don't depend on the general partner's word. When you buy into a private-company SPV, there is frequently no equivalent. The general partner tells you what the fund owns. The general partner's administrator confirms what the general partner says. There is no independent third party whose job is to verify the cap table entry exists and matches the fund's claimed position. OpenAI and Anthropic have both stated publicly that they don't authorize SPVs to hold their stock at all and that any unapproved transfer is void, which means even a scrupulously honest SPV sponsor can't always prove clean title, let alone a dishonest one (Cryptopolitan).

The third is timing pressure disguised as opportunity. Adit told investors they had to move fast to catch a pre-IPO window. That urgency is not unique to Adit. It's the standard pitch across the entire SPV market right now, because the underlying asset (a slice of a company that isn't selling shares to the public) is scarce by design, and scarcity plus urgency is what stops investors from asking the boring questions that would have caught this two years earlier.

The FOMO Trap That Makes This Work

I think the most underappreciated dynamic here is that the same hype cycle that makes an OpenAI or Anthropic SPV attractive is the mechanism that lets a fraudulent sponsor charge whatever markup they want without anyone questioning it. DeFiLlama reportedly values Anthropic at roughly $1.38 trillion and OpenAI at about $900 billion among the private companies it tracks, and demand for any sliver of exposure to either name is intense enough that sponsors can, and often do, layer fees, markups, and "acquisition costs" on top of the actual share price without investors pushing back (Cryptopolitan). Ask a hard question about fee structure or proof of ownership, and the implicit answer from the sponsor is that there are ten other investors waiting to take your allocation. Nobody wants to be the person who lost their spot in an OpenAI-adjacent fund over what feels like a compliance technicality. That fear is the entire business model.

Adit's own numbers show this isn't hypothetical: overcharging client funds with undisclosed "Acquisition Fees" the fund agreements didn't even permit, and reporting a misleading "Original Purchase Price" that concealed the markup, according to the SEC's complaint. Nobody caught it for years, across more than 1,000 investors, because nobody was positioned to ask.

The SEC's Ownership Push Is a Rearview Mirror

Since late August, the SEC has reportedly stepped up examinations of SPVs tied to registered investment advisers, asking firms to produce records proving their vehicles genuinely hold, or are exposed to, the private shares they market, according to the Wall Street Journal's reporting as relayed by multiple outlets (Quartz). These examinations reportedly involve document requests and sometimes in-person components, and can run from several weeks to close to a year. In my opinion, this is a useful step, and it's also fundamentally reactive. An examination cycle measured in months means the SEC is checking whether an SPV holds what it claims well after that SPV has already sold interests to investors. If a sponsor is running the Adit playbook today, that sponsor has months, possibly a year, of runway before an examination even starts asking the right questions, and the examination only reaches advisers registered with the SEC in the first place. The complaint notes Adit didn't register until March 2024, having relied on a venture-capital exemption the SEC says it didn't qualify for; a sponsor willing to skip registration or structure around SEC jurisdiction sidesteps this entire verification push.

The Linqto case makes the point concretely. That firm marketed pre-IPO stakes in Ripple, SpaceX, and Anthropic to smaller investors. An internal probe reportedly found customers never actually owned the underlying securities, and the firm has since filed for bankruptcy with the SEC investigating (Quartz). By the time that surfaced, the money was already spent, the company was already insolvent, and investors were left arguing over what, if anything, they held. Verification after the fact doesn't return capital. It documents the loss.

My Risk Framework for Any SPV

This next part is entirely my own view, built from watching this pattern play out, and I want to be clear it is not a guarantee against fraud. No checklist catches a determined liar. But it filters out most of the structural risk before you write a check.

Direct cap table confirmation. I want to see, or have my counsel see, documentation tying the SPV directly to a line on the target company's capitalization table or a transfer agent record, not a general partner's assertion that "the fund holds an interest." If the sponsor can't produce that, or says it's "in process," that's my answer.

Single-layer structure only. I will not invest in a fund that invests in a fund that invests in the shares. Every additional layer is a general partner I don't know, charging a fee I probably can't see, standing between my capital and the asset. One SPV, one direct holding, full stop.

A named, independent custodian or administrator. Someone other than the general partner needs to be responsible for confirming the position exists and matches what investors are told. If the fund's administrator is an affiliate of the general partner, that's not independence, that's the same conflict Adit exploited when it sold shares to its own client funds without disclosure.

No markup above documented cost, or full disclosed math. If a sponsor bought shares at $420 and is selling my allocation at $498, I want that spread stated in writing, in dollars, before I wire money, along with the sponsor's compensation on the deal. Silence on markup is the single clearest signal in the Adit case, and it's the easiest thing for a legitimate sponsor to simply tell you upfront.

I'd add one more filter that isn't in most frameworks: ask what happens to your capital if the company (OpenAI, Anthropic, whoever) never approves the transfer at all. Both companies have said they don't recognize unauthorized SPV positions in their stock. If the sponsor doesn't have a clean answer for that scenario, the allocation you think you're buying may carry no economic value regardless of who signed the wire instructions.

For more on this, see our coverage of SEC vs. Adit Ventures: A Pre-IPO SPV Fraud Case Study for Accredited Investors and 8-Step Checklist to Vet a Pre-IPO SPV Sponsor Before Committing Capital.

Frequently Asked Questions

Did Eric Munson admit to the SEC's allegations against Adit Ventures?

No. Munson and the other defendants consented to a judgment without admitting or denying the allegations, and Munson stated publicly that he settled to avoid the cost of prolonged litigation, not as a concession the charges have merit.

Are OpenAI and Anthropic shares actually being sold through fraudulent SPVs right now?

The Adit case centered on SpaceX, Klarna, and Flexport, not OpenAI or Anthropic directly, but both AI labs have separately warned that they don't authorize SPVs to hold their stock and that unapproved transfers carry no economic value, which is the same underlying ownership-verification problem the SEC is now examining across the SPV market.

What is the SEC actually asking investment firms to produce in these ownership-verification examinations?

According to Wall Street Journal reporting, examiners are asking registered investment advisers for documentation proving their SPVs genuinely own, or have real economic exposure to, the private-company shares marketed to investors, through document requests and sometimes in-person interviews that can run from several weeks to nearly a year.

Can an SPV investor lose everything even if the sponsor isn't intentionally committing fraud?

Yes, in my view this is underappreciated. If the underlying company never approves the share transfer, as OpenAI and Anthropic have both warned can happen with unauthorized SPV interests, an investor can hold a legally void position even when the sponsor believed in good faith that the transfer was valid.

Author Disclosure: Jeff Barnes, MBA has no personal position in any company, fund, or platform named in this article. Angel Investors Network has no current commercial relationship with any party mentioned. AIN provides marketing and education services, not investment advice. Past performance does not guarantee future results. All investments involve risk, including loss of principal.