OpenAI Warns of Mixpanel Data Breach Impacting API Users

OpenAI has informed its application programming interface (API) users that some of their data may have been exposed in a breach impacting Mixpanel, a data analytics supplier.

In a November 26 blog post , the generative AI company said an attacker gained unauthorized access to part of Mixpanel’s systems and exported a dataset containing limited customer identifiable information and analytics information. This incident started on November 9 and Mixpanel shared the dataset with OpenAI on November 25, following an internal investigation.

OpenAI warned that users of platform.openai.com, its API, may have been included in data exported from Mixpanel.

Affected data include:

OpenAI said its other products, including ChatGPT, and the underlying chat content, prompts, responses and API usage data were not impacted.

“This was not a breach of OpenAI’s systems. No chat, API requests, API usage data, passwords, credentials, API keys, payment details or government IDs were compromised or exposed,” it added.

The generative AI company removed Mixpanel from its production services and is supporting the data analytics provider in conducting a security investigation.

It has also started notifying potentially affected users and organizations.

“Beyond Mixpanel, we are conducting additional and expanded security reviews across our vendor ecosystem and are elevating security requirements for all partners and vendors,” OpenAI said in its blog post.

Mixpanel is an analytics platform that tracks user behavior inside applications and websites. OpenAI said it used its services “to help us understand product usage and improve our services for our API product.”

In a separate report published on November 27, OX Security outlined the type of information that can be collected via Mixpanel. These typically include:

However, the security company noted that most websites have their own custom analytics configurations, meaning the data collected by Mixpanel can vary.

OpenAI said the most likely use of this compromised data is through phishing or social engineering attacks. It urged people who think their API data may have been compromised to remain vigilant for credible-looking phishing attempts or spam and to implement the following best practices: