OpenAI employees have raised concerns about operational security gaps inside the company for years, according to Reuters, which cited sources familiar with the matter. The reporting, published Tuesday, indicates those internal warnings predate the security changes the company has since put in place.
Reuters reports that OpenAI has moved to strengthen its security operations following that history of internal alerts. The sources describe a pattern in which staff identified weaknesses well before the company acted on them.
The Reuters account arrives alongside other recent developments involving OpenAI personnel. A safety employee resigned, stating that the period for trial and error had ended, according to ETHRWorld.com. Separately, the San Francisco Chronicle reported that a grand jury indicted an OpenAI engineer accused of using toilet paper to bug a conference room. Coverage from Traders Union and Unite.AI also focused on the company's security posture and the years of internal warnings.
The details available so far are limited to the headlines and summaries of these reports. Reuters has not disclosed in the material available which specific lapses were flagged, when they occurred, or how the company responded at the time. The identities of the sources and the full scope of their accounts are likewise not described.
Taken together, the reports place OpenAI's internal security practices under scrutiny at a moment when the company is publicly adjusting them.
What to watch next: whether OpenAI comments publicly on the reported warnings and outlines what its tightened security operations now cover.
