(Credit: Chip Somodevilla/Getty Images)
OpenAI has appointed a former director of the National Security Agency (NSA) to its board, and that's not sitting well with Edward Snowden, the whistleblower who leaked the NSA's surveillance secrets a decade ago and now lives in Russia.
"Do not ever trust OpenAI," Snowden tweeted today after the company named retired US Army General Paul Nakasone to the board's new Safety and Security Committee.
OpenAI cites Nakasone's "world-class" cybersecurity experience; from 2018 to February 2024, he led both the NSA and US Cyber Command. The company is now hoping Nakasone will help OpenAI improve its cybersecurity, including protecting its AI supercomputers from hacking groups.
However, Snowden is alleging OpenAI has ulterior motives with the Nakasone hiring. "There is only one reason for appointing an NSA Director to your board. This is a willful, calculated betrayal of the rights of every person on Earth. You have been warned," tweeted Snowden.
Snowden didn’t elaborate, but OpenAI has attracted scrutiny for how it collects data to train its AI models. ChatGPT is processing chats from tens of millions of users per day — which could also make the technology a valuable way to conduct surveillance.
"I do think that the biggest application of AI is going to be mass population surveillance, so bringing the former head of the NSA into OpenAI has some solid logic behind it," tweeted security expert Matthew Green.
OpenAI didn’t immediately respond to a request for comment. But the company has treated Nakasone’s hiring as all about cybersecurity and responsible AI development. “The security of OpenAI’s systems—from protecting the large AI training supercomputers we operate to securing our sensitive model weights and the data entrusted to us by customers—is central to achieving our mission,” the San Francisco lab said in its announcement.
OpenAI also notes that "AI has the potential to deliver significant benefits...for many institutions frequently targeted by cyberattacks like hospitals, schools, and financial institutions."
Like other tech providers, the company can be compelled to disclose user information to comply with law enforcement requests, including a user’s name, contact details, IP address, and content. However, OpenAI does offer data controls that can block their data from being harnessed for AI model training purposes.