OpenAI Voids Employee NDAs After Multiple Former Staff Complaints

Key Takeaways

• One or more OpenAI employees accused the company of suppressing whistleblowers in a letter to SEC Chairman Gary Gensler.
• The letter said the company made staff sign non-disclosure agreements (NDAs) that violate SEC Rule 21F-17.
• However, CEO Sam Altman said OpenAI has now voided non-disparagement terms in its contracts.

OpenAI CEO Sam Altman has responded to accusations that the company silenced whistleblowers with restrictive (and as some employees alleged, illegal) non-disclosure agreements (NDAs).

Announcing that the company had voided non-disparagement terms for current and former employees in a post on X, Altman said : “we want  […] employees to be able to raise concerns and feel comfortable doing so.”

Gary Gensler Sent Urgent Letter About OpenAI NDAs That ‘Silence Employees’ 

Altman’s intervention comes after anonymous whistleblowers sent a letter to Securities and Exchange Commission (SEC) Chairman Garry Gensler voicing concerns about OpenAI’s NDAs.

In the letter, one or more whistleblowers accused OpenAI of violating SEC Rule 21F-17 (a), which protects whistleblowers who have reported possible securities law violations from retaliation by their employer.

“The agreements prohibited and discouraged both employees and investors from communicating with the SEC concerning securities violations, forced employees to waive their rights to whistleblower incentives and compensation, and required employees to notify the company of communication with government regulators.”

Alleged violations documented by the whistleblower(s) include: 

• Non-disparagement clauses that failed to exempt disclosures of securities violations to the SEC.
• Requiring prior consent from the company to disclose confidential information to federal authorities.
• Confidentiality requirements with respect to agreements, that themselves contain securities violations.
• Requiring employees to waive compensation that was intended by Congress to incentivize reporting and provide financial relief to whistleblowers.

The letter, which was first published by the Washington Post , called for the SEC to take “swift and aggressive steps” to enforce Rule 21F-17 within the AI sector and to hold OpenAI accountable for any violations.

Altman: ‘We’ve Worked Hard to Make it Right’

Without directly acknowledging the SEC whistleblower claims, Altman’s post on Thursday, Aug.1 attends to two of their accusations.

He said OpenAI has voided non-disparagement terms and contract provisions that gave it the right to cancel employees’ vested equity. “We’ve worked hard to make it right,” he concluded.

However, it isn’t clear whether the changes will be enough to appease the authors of the anonymous letter or a potential SEC investigation.

Enforcement Accelerates

Since adopting Rule 21F-17 in response to the 2010 Dodd-Frank Act, the SEC has brought over 20 cases against companies judged to have impeded whistleblower protections and incentives.

In recent years the agency has ramped up its enforcement action, with Chairman Gensler overseeing more Rule 21F-17 investigations than any of his predecessors. 

As well as accelerating the pace of enforcement, penalties for  21F-17 violations have also grown since the first one was issued in 2015.

In the largest fine to date, last year, the New York-based investment adviser D. E. Shaw agreed  to pay $10 million to settle the SEC’s charges that its employee agreements did not include sufficient whistleblower protection language.

Whistleblower Protections Critical for AI Safety

Although the anonymous employee(s) argued that OpenAI’s NDAs are in violation of the SEC’s rules, they situated their complaint within the wider protections established by the Dodd-Frank Act . 

Given the risks involved with the technology being developed, the letter argued that OpenAI’s overreaching NDAs are “particularly egregious” and called for the SEC to “send the message to others in the AI space that violations on the right of employees or investors to report wrongdoing will not be tolerated.”

The latest incident comes after 11 current and former OpenAI employees signed an open letter calling for increased protections for whistleblowers and a more transparent approach to AI safety.

Meanwhile, some OpenAI staff have criticized the firm for not taking AI safety seriously enough.

OpenAI Under Fire Over AI Safety 

In May, AI alignment researcher Jan Leike resigned from OpenAI over his concern that safety has taken a backseat to commercial interests. Leike has since taken up a post at rival AI developer Anthropic, which positions itself as a more responsible alternative to OpenAI.

Although the Dodd-Frank Act was designed with capital markets in mind, legal protections for whistleblowers have implications far beyond the financial sector.

From Facebook  to Boing, whistleblowers have been invaluable in exposing corporate misconduct. 

With AI rapidly extending its reach into all corners of American life, whistleblower protections are critical to ensure the technology is applied fairly and safely. If OpenAI’s NDAs are found to violate such protections, it could have consequences for the entire industry.