OpenAI, the artificial intelligence company co-founded by Sam Altman, is facing allegations from whistleblowers who have filed a complaint with the Securities and Exchange Commission (SEC). The complaint accuses OpenAI of unlawfully preventing its employees from raising concerns about the potential risks posed by its technology to regulators. The whistleblowers claim that the company’s employment, severance, and nondisclosure agreements were overly restrictive, potentially subjecting employees to penalties for reporting concerns to federal authorities.
According to a seven-page letter obtained exclusively by The Washington Post, OpenAI required its staff to waive their federal rights to whistleblower compensation and seek prior consent from the company before disclosing information to federal regulators. The letter argues that these agreements violated federal laws designed to protect whistleblowers who wish to disclose critical information anonymously and without fear of retaliation. The whistleblowers assert that such agreements send a message that the company discourages employees from engaging with federal regulators, hindering the scrutiny and dissent necessary for the development of safe and socially beneficial AI technology.
OpenAI spokesperson Hannah Wong responded to the allegations, stating that the company’s whistleblower policy safeguards employees’ rights to make protected disclosures. Wong also mentioned that OpenAI has already made changes to its departure process, removing nondisparagement terms. However, concerns persist that OpenAI’s focus on profit may be prioritized over safety in the development of its technology.
The recent complaint comes on the heels of reports that OpenAI rushed the release of its latest AI model, ChatGPT, despite employee concerns that the company failed to adhere to its own security testing protocol. The model’s release raised worries about potential catastrophic harms, such as the AI teaching users to build bioweapons or aiding hackers in developing new cyberattacks.
Confidentiality agreements in the tech industry have long been criticized for limiting employees’ ability to report misconduct, including sexual harassment and racial discrimination. Regulators have also expressed concerns that such agreements muzzle tech employees who could potentially expose misconduct related to algorithms that undermine elections, public health, and children’s safety. Policymakers have called for increased regulation of the AI industry, highlighting the need for whistleblowers to shed light on potential threats posed by rapidly advancing technology.
Senator Chuck Grassley emphasized the importance of addressing OpenAI’s restrictive agreements, stating that they appear to hinder whistleblowers’ rights and impede the federal government’s ability to stay ahead of AI developments. The letter sent to the SEC chairman, Gary Gensler, was also shared with Congress, and the SEC has acknowledged the complaint, although it remains unclear whether an investigation has been launched.
Stephen Kohn, the lawyer representing the OpenAI whistleblowers, argued that the agreements threatened employees with criminal prosecution under trade secret laws if they reported violations of the law to federal authorities. Kohn stressed the need for employees to come forward and for OpenAI to foster an environment of openness.
The letter calls on the SEC to take swift action against OpenAI, requesting the examination of all employment, severance, and investor agreements containing nondisclosure clauses to ensure compliance with federal laws. The whistleblowers also urge federal regulators to notify past and current employees of the violations committed by OpenAI and their right to confidentially and anonymously report any legal violations to the SEC. The letter suggests that fines should be imposed on OpenAI for each improper agreement, and the company should be directed to rectify the chilling effect caused by its past practices.
The battle against Silicon Valley’s use of nondisclosure agreements to control information has been ongoing. Recently, lawyer Chris Baker secured a $27 million settlement for Google employees, who alleged that the company employed onerous confidentiality agreements to suppress whistleblowing and other protected activities. Baker noted that tech companies are increasingly finding ways to deter speech, recognizing that the cost of leaks can outweigh the cost of litigation.