Hi, everyone – welcome to BLPB at the new place! As you can see, we’ve imported our old posts over here but in the process, the authors got scrambled for a lot of the older ones … we’ll work that out eventually. So! Moving on –

OpenAI’s been in the news, again, and there are some interesting things to talk about.

First, apparently both OpenAI, and Anthropic before it, have been raising funds through SPVs.  The SPV formally invests in the company; the SPV raises the funds from the real investors.

The purpose of that is the securities laws. If OpenAI/Anthropic have too many investors, they’ll have to make public filings.  But each SPV counts as a single investor.  So, by taking investment through SPVs, OpenAI and Anthropic can keep their formal investor count below the threshold of becoming a public company, while in fact raising capital from a dispersed group.

That said, it only works if OpenAI/Anthropic are not the ones organizing the SPVs. If they are, they’re evading the statutory limit on the number of investors a private company can take on.  The SEC’s view is that this trick only works if the SPVs are organized independently of the issuer.  I had a blog post on this way back when it was relevant to Uber.

When Uber did this, the impression I got was that investment banks created the SPVs to please clients who wanted access to a hot startup, and who were wealthy but didn’t, individually, have the kind of cash that attracts the attention of venture-backed companies.  But OpenAI’s SPVs are not being organized by investment banks; they’re being organized by large venture capital firms, who have preexisting investments in OpenAI. 

According to Anat Alon-Beck and John Livingston, this is becoming increasingly common, suggesting we’ve come a long way from investment banks trying to please clients who want access, and now this is just an alternative way to meet the capital raising needs of startup companies while allowing them to stay private.  OpenAI, we know, has enormous need for capital.

Point being, I wonder in these situations just how “independent” the SPV capital raises are, which is supposed to be a precondition to avoid counting their investors as investors in the issuer.  After all, VC firms often get governance rights along with their investments, they provide counseling/guidance to their portfolio firms, so in some sense they “are” the issuer, potentially going out and organizing an SPV so that they can funnel money to the firm while staying under the securities laws’ radar.  That whole process not only exposes perhaps smaller investors to increased risks, but also permits large and societally important firms to stay dark, without exposing their operations to public scrutiny.

But that’s not the only interesting OpenAI news!  There are also reports that it may abandon its current structure – an LLC, where investors can receive only a capped profit, operated by a nonprofit – in favor of a public benefit corporation.

That fact itself tells you something about the toothlessness of the public benefit corporation structure, but one of the weirder things is that apparently they’re claiming the public benefit form will help fend off activist attacks.

As I’ve previously discussed, there’s nothing about the benefit corporation form that, in fact, does prevent activism. OpenAI is still a private company, and I imagine is quite a long ways from going public, which means it’s insulated from activist attacks currently.  Even if it does go public, I find it difficult to imagine it wouldn’t have a multi-class share structure that would be sufficient to fend off activists, and since the benefit corporation form itself doesn’t do that work, I view the choice of the benefit corporation form as greenwashing, i.e., an attempt to put a happier face on the fact that the company is abandoning the nonprofit mission that was supposed to protect humanity, or whatever.

But maybe I’m wrong about that.  This is an interesting question, we could ask, post Coster v. UIP Cos. In Coster, the Delaware Supreme Court laid out the framework for assessing defenses that interfere with shareholder voting:

When a stockholder challenges board action that interferes with the election of directors or a stockholder vote in a contest for corporate control, the board bears the burden of  proof. First, the court should review whether the board faced a threat “to an important corporate interest or to the achievement of a significant corporate benefit.” The threat must be real and not pretextual, and the board’s motivations must be proper and not selfish or disloyal. As Chancellor Allen stated long ago, the threat cannot be justified on the grounds that the board knows what is in the best interests of the stockholders.

Second, the court should review whether the board’s response to the threat was reasonable in relation to the threat posed and was not preclusive or coercive to the stockholder franchise. To guard against unwarranted interference with corporate elections or stockholder votes in contests for corporate control, a board that is properly motivated and has identified a legitimate threat must tailor its response to only what is necessary to counter the threat. The board’s response to the threat cannot deprive the stockholders of a vote or coerce the stockholders to vote a particular way.

Is there an argument that benefit corporation directors have more leeway to interfere with shareholder voting, say, if an activist wants to run a proxy contest, than they would in an ordinary corporation?  After all, Coster says a defense cannot be motivated by the directors’ belief they know what is best for stockholders – but what if the defense were motivated by the directors’ belief that the stockholders’ interests must not be permitted to overwhelm a stakeholder’s interest in a corporation established to advance that interest?

My suspicion is, it shouldn’t matter.  The form itself represents a choice by investors to pursue a social mission; which means that ultimately the mission rests in their hands.  Plus, Delaware amended its statute to permit conversion to, and away from, the benefit corporation form with a mere 50-percent vote (rather than a supermajority), and also removed statutory provisions that would have granted shareholders in ordinary corporations appraisal rights upon conversion to a benefit corporation.  All of which suggests that the principle that the “shareholder franchise is the ideological underpinning upon which the legitimacy of directorial power rests” remains just as intact in a benefit corporation as it does with any other corporation, which means there should be no special license to interfere with shareholder voting.  I mean, if shareholders can convert to ordinary corporation status with a mere 50% vote, surely directors can’t interfere with shareholders’ choice to do that kind of thing.

Plus, given how weak the enforcement rights are in benefit corporations alleged to have strayed from their social missions, I suspect that if Delaware courts were to rule that the form grants directors more leeway to fend off activists, suddenly corporate America would discover a new love of social purpose.

But, I suppose this is an issue lurking around that Delaware may one day have to confront.

And another thing.  New Shareholder Primacy podcast is up!  This time, me and Mike Levin talk SEC climate change rules, and the activist attack at Pfizer. Available at Apple, Spotify, and YouTube.