Should We Care Whether OpenAI Employees — not just Board Members — Are Disqualified Persons?
When I was practicing higher education law at a certain University that shall never ever ever again have my loyalty and support, the General Counsel occasionally engaged [expensive] outside counsel to investigate media allegations that some young superstar athlete was seen riding around in a fancy car with a known sports agent. That sort of thing was frowned upon because schools could lose lots of TV and bowl game revenue for violating NCAA indentured servitude rules. I never really understood the purpose of the outside counsel exercise. Its like a criminal suspect hiring his own polygrapher to administer a test and then pronounce the suspect not guilty. Isn’t that conclusion an implicit requirement of the engagement? It seems easily contrived and just as easily impeached.
Still OpenAI did just that, it hired a firm to investigate its governance and pronounce its governance nonprofit. And on Friday it pronounced itself not guilty, still nonprofit, and deserving of tax exemption:
The OpenAI Board today announced the completion of the review by WilmerHale. The firm conducted dozens of interviews with members of OpenAI’s prior Board, OpenAI executives, advisors to the prior Board, and other pertinent witnesses; reviewed more than 30,000 documents; and evaluated various corporate actions. Based on the record developed by WilmerHale and following the recommendation of the Special Committee, the Board expressed its full confidence in Mr. Sam Altman and Mr. Greg Brockman’s ongoing leadership of OpenAI. “We have unanimously concluded that Sam and Greg are the right leaders for OpenAI,” stated Bret Taylor, Chair of the OpenAI Board.
What else would any law firm conclude? OpenAI is a client worth $90 billion and still counting. Ain’t nobody trying to bite the hand that feeds. OpenAI’s blog post pronouncing itself objectively proven innocent also listed three new independent directors:
- Dr. Sue Desmond-Hellmann, former CEO of the Bill and Melinda Gates Foundation and on the Board of Directors at Pfizer and on the President’s Council of Advisors on Science and Technology.
- Nicole Seligman, former EVP and Global General Counsel of Sony and President of Sony Entertainment and on the Board of Directors at Paramount Global, Meira GTx, and Intuitive Machines, Inc.
- Fidji Simo, CEO and Chair of Instacart and on the Board of Directors at Shopify
I am not here to question any board member’s bona fides. But a few weeks ago, me and Ben Leff spoke to David Brennen’s Nonprofit class at the aforementioned and now despised university’s LLM program. We agreed that OpenAI is pretty much doing the joint venture with Microsoft in accordance with Rev. Rul. 98-15. But I thought that the fact that some or all of its insiders had stock options in the profit side of the venture ought to be fatal. By insiders I meant OpenAI board members but Ben informed me that none of the board members have stock in the profit side. He cited Aprill, Loui, and Horwitz and sure enough that’s what they report. Only OpenAI researcher/employees have stock options worth millions, if not billions by now.
So why am I still bothered by the manner in which Microsoft [apparently] flexed its profit-making muscle to reinstall Altman to the Board? Here is my charge: Indispensable OpenAI employees have stock options in the $90 billion profit side sufficient to negate OpenAI’s 98-15 governing authority over the joint venture; or if not that, the employees might be engaging in transactions thought ought to be tested by private inurement and excess benefit analysis. If either charge is proven, OpenAI would not be entitled to exemption. I’m not advocating that result. I want OpenAI to win the case for exemption and am just pointing out factual and theoretical weaknesses.
Here is my specification: joint venture employees — all of whom were or are hired to perform OpenAI’s tax exempt charitable research, but not serve on the board — have financial interests in the joint venture’s profit side. They are also indispensable to the charitable non-profit tax exempt effort. They control everything about the operation and yet they share in the profits. Does that not bother anybody? Essential [and controlling] employees have a financial stake in the for profit partner. Normatively, we ought to ensure that in for-profit nonprofit joint ventures, those in control are conflict free. Those in practical control, not just officers and directors. The doctrinal point is that you don’t have to be a an officer or board member to be an insider or disqualified person.
Here is what Treas. Reg. 53.4966-3(e)(2) says about whether employees who are neither officers or directors may still have “substantial influence:”
(2) Facts and circumstances tending to show substantial influence. Facts and circumstances tending to show that a person has substantial influence over the affairs of an organization include, but are not limited to, the following—
. . .
(iii) The person‘s compensation is primarily based on revenues derived from activities of the organization, or of a particular department or function of the organization, that the person controls;
(iv) The person has or shares authority to control or determine a substantial portion of the organization’s capital expenditures, operating budget, or compensation for employees;
(v) The person manages a discrete segment or activity of the organization that represents a substantial portion of the activities, assets, income, or expenses of the organization, as compared to the organization as a whole;
. . .
I’d say OpenAI research employees check all three of thise boxes. So look. The employees shouldn’t be excluded from the suspect list now that they are invested in and stand to profit immensely from the for-profit side. And it was reportedly the employees with equity interests in, and entire control over the research who engineered Altman’s return to the Board, and then the ouster of Board members who had different ideas than Altman. Altman’s ideas are reportedly more to Microsoft’s liking and Microsoft comprises the profit side of the joint venture. By its financial relationship with those employees, Microsoft exercised control over at least the selection of board members. Those absolutely indispensable employees — they personify everything that OpenAI is — threatened to leave (and go work for the for-profit partner even!) if Altman was not reinstated. Ben, Ellen, Rose and Jill seem content that the OpenAI meets all that is required by Rev. Rul 98-15 at the board level and sure enough it does. But somehow I think they implicitly define insiders and disqualified persons too narrowly if 4958 is to achieve its purpose of policing conflicts of interests harbored by high level fiduciaries.
So yeah, OpenAI is clearly structured to find refuge in the 98-15 safe harbor. Harbors ought not be impregnable. If we see something inside the harbor that doesn’t look right we should have the right to go in and check it out. OpenAI maintains majority control of the joint venture on paper, but if invested employees are disqualified persons or exercise practical control despite the formal corporate structure, we ought to inquire. Its actual control we are worried about when we talk about joint ventures. So that charity will prevail over profit. And we should at least inquire about private inurement and excess benefit whenever an indispensable or controlling employee has divided interests. Like the OpenAI researchers and their $90 million stock options.
darryll k. jones