Leaked Docs Expose OpenAI’s Hardball with Ex-Staffs, Sam Altman Apologized

According to the reports, employees at the prominent tech company OpenAI faced extensive and very restrictive exit agreements if they decided to depart from the organization.

Those who declined to sign these documents promptly were allegedly warned about the potential loss of their vested equity in the company, a measure that is not commonly seen in Silicon Valley.

This policy resulted in former employees having to make a difficult decision between forfeiting potentially substantial earnings they had already accumulated or accepting a clause that prevented them from criticizing the company indefinitely.

Insiders from within the company revealed that the news sparked a major uproar at OpenAI, a privately held company currently estimated at around $80 billion.

Similar to numerous startups in Silicon Valley, staff members at OpenAI typically receive a significant portion of their total anticipated remuneration in the form of company shares.

There is a common belief among employees that once these shares have “vested” based on the agreed schedule in their contract, they become their property and cannot be revoked, just like a company wouldn’t retract already disbursed salary.

Leaked Docs Expose OpenAIs Hardball with Ex Staff Sam Altman Apologized

CEO Sam Altman posted an apology, saying

We have never clawed back anyone’s vested equity, nor will we do that if people do not sign a separation agreement (or don’t agree to a non-disparagement agreement). vested equity is vested equity, full stop.

There was a provision about potential equity cancellation in our previous exit docs; although we never clawed anything back, it should never have been something we had in any documents or communication. this is on me and one of the few times i’ve been genuinely embarrassed running openai; i did not know this was happening and i should have

Some members of OpenAI’s executive team have also expressed their apologies in internal communications.

OpenAI’s chief strategy officer, Jason Kwon, admitted in a leaked message to employees that although the provision had been in effect since 2019, it was only noticed by the team approximately a month ago.

Kwon took responsibility by stating, “The fact that it went unnoticed for this long is my responsibility.”

However, there is an issue with the apologies issued by the company’s leadership. Documents from the company, including signatures from Altman and Kwon, challenge their assertion that they were unaware of the clawback provisions.

A termination letter found in the documents states clearly that individuals with vested Units must sign a release of claims agreement within 60 days to keep their Units.

This letter is signed by Kwon and Diane Yoon, the former OpenAI VP of people. Additionally, the highly restrictive NDA, signed solely for the consideration of vested equity, is endorsed by COO Brad Lightcap.

Documents provided to the reporters by former employees reveal that the incorporation documents for the holding company managing equity in OpenAI include several sections granting the company extensive power to reclaim equity from former employees or, crucially, prevent them from selling it.

Sam Altman, acting as the CEO of OpenAI, signed the incorporation documents on April 10, 2023. When asked by reporters about the clauses in the documents that were potentially unknown to Altman, OpenAI did not directly address the question.

However, Kwon expressed regret in a statement to reporters, saying, “We apologize for the concern this has raised among our dedicated team. Resolving this promptly is our priority, and we are committed to improving our practices.”

The apparent inconsistency between the recent statements from OpenAI’s leadership and the contents of these documents has implications that extend well beyond financial matters.

OpenAI stands out as one of the most influential companies in the field of artificial intelligence, being highly visible. The company has set a clear goal of “ensuring that artificial general intelligence is used for the benefit of all humanity.”

However, introducing artificial general intelligence to society is a responsibility that requires significant public trust and complete transparency.

If OpenAI employees don’t feel comfortable expressing criticism without fearing financial consequences, how can the company and its CEO expect to earn that trust?

Aggressive methods used by OpenAI

As you read the extensive leaked documents, a consistent direction becomes apparent. Persuading former employees to agree to the highly restrictive nondisparagement and nondisclosure terms included threats of revoking their equity – yet there were additional tactics at play.

We reviewed two cases where OpenAI sent out lengthy and complex termination documents that expired after seven days.

This put former employees under pressure to decide whether to accept OpenAI’s conditions within a week or potentially lose out on significant amounts of money.

The short timeframe made it challenging for them to seek advice from external legal counsel.

Ex-employees who requested additional time to seek legal advice and review the documents encountered strong resistance from OpenAI.

One employee, who sought an extra week to assess the intricate paperwork, was informed by a company representative via email this spring that the General Release and Separation Agreement must be signed within 7 days.

An OpenAI representative sent a follow-up email to a second employee who had requested an additional two weeks to review the agreement, emphasizing the potential impact on their equity if they choose not to sign.

The representative clarified that this policy applies universally and is being carried out according to established procedures.

Many former employees surrendered to the pressure, but for those who persevered, the company resorted to another strategy described by a former employee as the “legal retaliation toolbox” when they left.

Upon refusing to sign the initial termination agreement and seeking legal advice, the company shifted its approach. Instead of threatening to revoke his equity for not signing, they warned that he might be barred from selling his equity.

The subsequent documents that the company provided to him, which have been examined by reporters, state, “If you possess any vested Units and fail to sign the exit paperwork, such as the General Release, as mandated by company policy, it is crucial to comprehend that, among other consequences, you will lose the opportunity to take part in upcoming tender events or other liquidity opportunities organized or supported by our private company.” Essentially, signing is necessary to retain the option to sell your shares.

How OpenAI adopted a tough stance

In order to comprehend this and to grasp why OpenAI’s recent apology seems insincere, it’s essential to understand the concept of equity at OpenAI.

In a publicly traded company such as Google, equity simply refers to shares of stock. Employees receive a portion of their compensation in the form of salary and another portion in Google stock, which they can either retain or trade on the stock market like any other shareholder.

In private companies such as OpenAI, employees receive ownership shares of the company (or more commonly, options to buy ownership shares at discounted prices), but they must wait for a chance to sell these shares, which might not occur for several years.

Occasionally, large private companies conduct “tender offers,” allowing employees and ex-employees to sell their equity. OpenAI conducts tender offers periodically, with specific information closely guarded as a secret.

OpenAI’s announcement implies that individuals who choose not to sign the restrictive agreement will be excluded from participating in any upcoming tender offers.

This essentially links the equity, which is estimated to be worth millions of dollars, to the employee signing the agreement. This allows OpenAI to maintain that they have not revoked anyone’s vested equity, in line with Altman’s statement on May 18th.

The reporters contacted OpenAI to inquire about whether the organization has implemented or intends to implement a strategy that prevents former employees from accessing equity.

An OpenAI representative responded by stating, “In the past, ex-employees have had the opportunity to sell at a consistent price irrespective of their current employment status; we anticipate maintaining this approach.”

The individual responsible for informing a former employee that they would be ineligible for future tender offers without signing is unknown.

The former employees expressed concern that despite any public statements of reassurance the company might make, the legal documents OpenAI has in place provide numerous options for legal action, which could make it challenging for the company to back down from any particular issue.

Apart from the clauses stipulating that vested equity will be forfeited if a former employee fails to sign a general release within 60 days, the incorporation papers also include clauses indicating that, “at the company’s sole and complete discretion,” any terminated employee may have their vested equity holdings reduced to zero.

Additionally, there are provisions granting the company full discretion to decide which employees can take part in tender offers for the sale of their equity.

A source closely related to the company mentioned that the documents should prioritize the goal of creating a secure and advantageous AGI, but instead, they outline various methods to take action against employees who leave and express criticism towards the company.

Sam Altman signed these documents. OpenAI did not provide a response when asked if there was a discrepancy between Altman’s public statements claiming he was unaware of equity clawback language in company documents and the existence of such clauses in incorporation documents bearing his signature.

Stories You May Like

Help Someone By Sharing This Article