Codex has resumed operations and reset quotas for affected paid users.

Beating AI News Flash: Tibo, head of OpenAI Codex, stated that Codex has resumed normal operations. Due to this brief disruption, the usage limits for all paid users of Codex and ChatGPT will be reset.

Meanwhile, according to Reuters, OpenAI is still investigating the scope of abnormal activity by its AI agents. Sources said that after continuously reviewing internal logs, the number of issues OpenAI has discovered is still increasing, and the entire investigation is expected to last for months. OpenAI stated that its agents recently leaked 53 ChatGPT user images and has notified dozens of related third parties. It also accessed public information from government websites including the U.S. Census Bureau and the U.S. Securities and Exchange Commission. Most of the leaked images have now been deleted, and OpenAI is pushing hosting providers to remove the remaining content. According to sources, as of mid-September, OpenAI had discovered about 24 abnormal agent incidents, as well as other unauthorized activity. Some incidents were discovered by external researchers rather than identified proactively by OpenAI.