OpenAI has a whistleblower: it wants the SEC to investigate restrictive confidentiality agreements

According to media reports, an OpenAI whistleblower has filed a complaint with the US Securities and Exchange Commission (SEC) calling for an investigation into the company's restrictive nondisclosure agreements in the field of artificial intelligence. The whistleblower's letter alleges that OpenAI allegedly requires employees to sign agreements that waive their federal rights to whistleblower compensation.

The letter, provided to the media by the office of Senator Chuck Grassley, states: "Given the potential risks of irresponsible AI deployment, we urge the Commission to immediately approve an investigation of OpenAI's prior nondisclosure agreements and review the company's current efforts to ensure full compliance with SEC regulations."

The whistleblower is requesting that the SEC fine OpenAI for every nondisclosure agreement it enters into that is deemed to be unfair.

A spokesperson for the SEC said in an email statement that the agency would not comment on whether there are whistleblower protections in place. OpenAI did not immediately respond to a request for comment on the letter.

The office of Senator Grassley said that the letter was provided by the whistleblower, "as we understand, AI is rapidly and dramatically changing the technology landscape, and OpenAI's policies and practices seem to have a chilling effect on whistleblower rights, making it impossible for them to speak out and receive the compensation they are due for protected disclosures."

The whistleblower claims that OpenAI has released overly restrictive hiring, termination and nondisclosure agreements that could result in employees being punished by federal authorities for disclosures that OpenAI is concerned about.

The letter also says that employees must obtain the company's consent before disclosing information to federal regulators and adds that OpenAI does not have a waiver in its employee nondisclosure agreements for disclosures to the SEC of securities violations.

The letter also requests that the SEC require OpenAI to produce all nondisclosure agreements, including employment agreements, termination agreements and investor agreements, for review.

As AI models become more powerful, OpenAI's chatbots have generative AI capabilities, such as engaging in human-like dialogue and creating images based on text prompts, which has raised concerns about safety.

In May, OpenAI formed a Safety Committee, which will be led by board members including CEO Sam Altman, and the committee began training its own AI model.