He helped build ChatGPT. Then he decided it had been built by breaking the law and he was willing to say so under oath.
His name was Suchir Balaji. He was 26.
Born in 1998. Grew up in Cupertino, California, in the heart of Silicon Valley. A coder before most kids can drive. As a teenager he reached the finals of a major coding competition. He studied computer science at UC Berkeley and finished at 22.
Then, in 2020, he joined OpenAI early. Before ChatGPT. Before most people knew the name.
He helped build the thing. For about four years he worked on gathering the enormous piles of internet text used to train the models that became ChatGPT. The tool now used by hundreds of millions of people.
At first he didn't question it. He assumed the company could use whatever was on the internet. That was just how it worked.
Then ChatGPT exploded. And he started thinking hard about what it had actually been trained on.
Books. Articles. Websites. Real work by real people. Nobody asked them. Nobody paid them. The model absorbed all of it then turned around and competed with the very people whose work it had learned from.
He did the analysis himself and wrote it up. His conclusion: OpenAI's use of copyrighted data failed the legal test for fair use. He believed it broke copyright law.
He couldn't unsee it.
"If you believe what I believe," he said, "you have to just leave the company."
So he did. August 2024. He walked away from one of the most coveted jobs in tech reportedly paying around $350,000 a year.
Then he spoke.
October 2024. The New York Times published an interview with him. He posted his own essay laying the argument out in plain terms: OpenAI had trained ChatGPT on copyrighted work without permission, and he believed it was illegal.
And he went further. He said he was willing to testify. In the copyright lawsuits piling up against OpenAI, his name surfaced as someone holding relevant documents.
A 26-year-old, ready to take the stand against one of the most powerful companies on earth.
A month later, he was dead.
November 26, 2024. Suchir Balaji was found in his San Francisco apartment. He was 26 years old.
The San Francisco medical examiner ruled the death a suicide and reported no evidence of foul play. The apartment, investigators said, had been locked from the inside.
His family does not accept that ruling. Not for a moment.
His mother has called it cold-blooded murder. The family hired a former FBI agent as a private investigator. They paid for a second, independent autopsy. They point to things they say were out of place, and questions they believe the official investigation never answered. They have demanded the case be reopened. A member of Congress has backed calls for a federal investigation. Even Elon Musk weighed in, saying publicly that it didn't look like a suicide to him.
In September 2025, his parents filed a wrongful-death lawsuit — not against OpenAI, but against the owners of his apartment building, accusing them of concealing evidence and obstructing the investigation. As of 2026, that case is still moving through the courts.
No one has been charged. The official ruling still stands. And the family is still fighting to overturn it.
Here is what is not in dispute.
A 26-year-old helped build a technology used by hundreds of millions of people. He came to believe it had been built by breaking the law. He had a dream job, a brilliant future, and every reason to stay quiet and keep cashing the check.
Instead he put his real name on the accusation, walked away, and agreed to testify against a company worth billions.
And then, weeks before he could take the stand, he was gone.
The copyright war he warned about is still burning through the courts right now. His case became part of a national push to protect AI whistleblowers. And his parents are still standing outside OpenAI's offices, asking for answers the rest of us still don't have.
He was 26.
No comments:
Post a Comment