OpenAI whistleblower dies by suicide after questioning AI’s controversial practices

Suchir Balaji, a 26-year-old former OpenAI researcher-turned-whistleblower, was found dead in a San Francisco apartment last month. The San Francisco medical examiner’s office determined his death on November 26 as a suicide, with police finding no evidence of foul play.

Balaji, who left OpenAI in August, had recently spoken openly about the company's practice of training its chatbot on copyrighted material scraped from the internet. The artificial intelligence (AI) giant is currently facing multiple lawsuits over its data-gathering practices.

An Indian American raised in Cupertino, California, Balaji was exceptionally talented. He excelled in programming competitions, achieving 31st place in the ACM ICPC 2018 World Finals and securing first place in the 2017 Pacific Northwest Regional and Berkeley Programming Contests.

Balaji, like many in his field, was fascinated by artificial intelligence from an early age. In an October interview with The New York Times, he shared that his interest in AI began in his teenage years after reading a news article about the technology. He envisioned that neural networks could address humanity’s most significant challenges. 

"I thought that AI was a thing that could be used to solve unsolvable problems, like curing diseases and stopping ageing...I thought we could invent some kind of scientist that could help solve them," he said, according to the NYT report.

In 2020, he joined OpenAI, part of a wave of Berkeley graduates entering the company.

Balaji worked at OpenAI for four years. For one and a half of those years, he focused on collecting and organizing the vast amounts of internet data used to train the company’s chatbot, ChatGPT.

In the NYT interview, Balaji admitted that during his early days at OpenAI, he did not deeply consider the legality of using both copyrighted and open internet data to develop the company’s products. It was only after ChatGPT’s release in late 2022 that he began questioning the ethics of these practices, realizing that tools like ChatGPT were harming the internet by utilizing copyrighted material without authorization.

By 2024, Balaji concluded that "he no longer wanted to contribute to technologies that he believed would bring society more harm than benefit." He resigned from OpenAI in August without securing another job, instead focusing on "personal projects."

He passed away a day after being named in a court filing as someone whose records OpenAI would examine as part of a lawsuit filed against the company.