Senators pressure OpenAI over security concerns after whistleblower complaint

Several senators pressed OpenAI for answers on Monday about its security and employment practices, after a group of whistleblowers filed a complaint alleging the company prevented the team from warning regulators about the risks of its artificial intelligence (A.I.) technology. ).

Led by Sen. Brian Schatz (D-Hawaii), the group of mostly Democratic senators asked OpenAI CEO Sam Altman about the AI ​​startup’s public commitments to security, as well as its treatment of current and former employees expressing concerns.

“Given OpenAI’s position as a leading AI company, it is important that the public can trust the security of its systems,” Schatz, alongside Senators Ben Ray Lujan (D-N.M.), Peter Welch (D-Vt. ), Mark Warner (D-Va.) and Angus King (I-Maine), wrote in Monday letter.

“This includes the integrity of the company’s security testing and governance structure, its labor practices, its fidelity to its public promises and mission, and its cybersecurity policies,” they continued.

The startup behind popular AI chatbot tool ChatGPT is under increased scrutiny after the Washington Post obtained a complaint filed by multiple whistleblowers with the Securities and Exchange Commission (SEC) earlier this month.

The whistleblowers alleged that OpenAI granted its employees restrictive employment, severance and confidentiality agreements that required them to waive their federal rights to whistleblower compensation and penalized them for raising concerns with regulators.

“Given the risks associated with the advancement of AI, there is an urgent need to ensure that employees working on this technology understand that they can file complaints or address concerns with federal regulatory or law enforcement authorities,” the whistleblowers wrote in their complaint.

In Monday’s letter to Atlman, the senators asked OpenAI’s CEO to confirm that the company will not impose permanent non-disparagement agreements on its employees and to commit to removing any other provisions that could be used to penalize employees for speak publicly.

“If not, explain why and any internal protections in place to ensure these provisions are not used to financially disincentivize whistleblowers,” they added.

An OpenAI spokesperson told The Hill “artificial intelligence is a transformative new technology and we appreciate the importance it has for U.S. competitiveness and national security. We take our role in developing safe and secure AI very seriously and continue to work closely with policymakers to establish appropriate safeguards moving forward.”

OpenAI also made changes to its exit process to remove non-derogatory terms, according to the spokesperson.

Updated at 1:26 p.m. ET.

This story originally appeared on thehill.com read the full story