-
Share
OpenAI’s Security Blunders Raise Cause for Concern
An purported undisclosed data breach, unencrypted personal data, and the departure of two safety leads leaves OpenAI’s security culture in question.
OpenAI allegedly suffered a data breach last year, as threat actors hacked into the AI company’s internal messaging systems and stole information regarding its AI technology design, the New York Times reported.
Apparently occurring April of last year, OpenAI didn’t disclose the data breach to the general public as customer and partner data was unaffected by the breach. However, the tech company apparently went so far as to not inform law enforcement of the incident as they didn’t consider the breach to be a threat to national security and was instead committed by a private individual.
No systems were accessed – instead, details from an online forum where OpenAI employees discussed the latest technological incidents were allegedly breached.
The findings come as the ChatGPT app for Apple Macs was allegedly designed with little consideration for privacy.
Recommended reading
- Report: Vulnerability Exploitation Surge Endangers Cybersecurity
- IT Leaders Split on Generative AI’s Role in Cybersecurity
- Cyber Leaders Reveal Compliance and Boardroom Struggles
In a post on X, software developer Pedro José Pereira Vieito claimed that the ChatGPT app for macOS dismissed the inbuilt sandboxing the operating system uses to prevent private data exposure, instead storing user conversation in an unsecured and plain text records.
The OpenAI ChatGPT app on macOS is not sandboxed and stores all the conversations in **plain-text** in a non-protected location:
~/Library/Application Support/com.openai.chat/conversations-{uuid}/
So basically any other app / malware can read all your ChatGPT conversations: pic.twitter.com/IqtNUOSql7
— Pedro José Pereira Vieito (@pvieito) July 2, 2024
While OpenAI updated its sytem and encrypted these locally-stored directories, departures of high-ranking employees points to a potential crisis in the company’s security policies.
Chief scientist Ilya Sutskever, and head of the company’s previous safety team, Jan Leike, both left the company, with many tying the exits to concerns over OpenAI’s overarching safety culture.
