OpenAI: Personal data of some users exposed after security incident

(Patria) - Global AI company OpenAI has confirmed that the personal data of some users of its API service was exposed after a third party, the analytical service Mixpanel, suffered a security incident.

According to an official statement, an attacker gained unauthorized access to Mixpanel's systems and exported a dataset containing limited but sensitive information: user name, email address, approximate location (city, state, country), as well as data about the operating system, browser, web referrers, and organization or user ID within the API account.

It is important to emphasize that the incident did not affect OpenAI's systems themselves – chat content, API keys, passwords, payment information, or other critical information were not compromised.

Immediately after discovering the problem, OpenAI discontinued the use of Mixpanel for its production service and launched an internal investigation. The company also began the process of notifying affected users and companies, while raising security standards for all future engagements with third parties.

As a precaution, OpenAI urges all API service users to be vigilant about suspicious emails or messages – it is possible that attackers may exploit the exposed information for phishing or social engineering.

It is recommended to enable multi-factor authentication (MFA) and not to disclose passwords, API keys, or verification codes via email, chat, or SMS.