According to CNBC, Business Insider and other foreign media reports, the former employee of the artificial intelligence giant OpenAI, Suchir Balaji, was found dead in his apartment in San Francisco at the end of November, at the age of 26, and the coroner confirmed that the cause of death was suicide.
The San Francisco Police Department stated that on the afternoon of November 26, officers were called to Balaji's apartment in San Francisco to conduct a "welfare check," and found that Balaji had already passed away, and in the preliminary investigation, they ruled out the possibility of homicide.
It is worth noting that Balaji, as a whistleblower, publicly accused OpenAI of violating US copyright law by collecting data from the internet to train its AI models when building ChatGPT in October, which attracted widespread media attention.
Former employee reported that OpenAI's training of ChatGPT violated copyright law
Previously, Balaji left OpenAI, where he had worked for 4 years, in the summer of this year, and published articles on his personal blog and social media, publicly questioning his former company's unethical use of internet data to train its AI models, and allegedly violating US copyright law.
In his blog post, he stated: "While generative AI models rarely produce outputs that are fundamentally similar to their training inputs, the process of training generative models does involve copying copyrighted data. If these copies are unauthorized, this could be considered copyright infringement, depending on whether the specific use of the model qualifies as 'fair use'. Since fair use is determined on a case-by-case basis, it is not possible to make broad statements about when generative AI would qualify for 'fair use' exemptions."
In his last tweet on October 24, he also said that he believes ChatGPT and similar chatbots will undermine the commercial value of the digital data and content currently widely used to train AI models, for both individuals and organizations that create them.
For many generative AI products, 'fair use' seems to be a rather incredible defense, the root cause of which is that they can create substitute products that compete with the data they were trained on.
OpenAI whistleblower committed suicide in his apartment
OpenAI is facing multiple lawsuits accusing the company of copyright infringement, including from newspapers and media publishers such as The New York Times and The Times of London, as well as from various creators and artists. However, in recent months, OpenAI has signed commercial agreements with media such as Time Magazine and the Financial Times, allowing the company to use the content of these publishers to train its AI models.
Previously, in response to Balaji's accusations, OpenAI issued a statement saying: "We use publicly available data to build AI models, in a way that is supported by the principles of fair use and relevant precedents that have been widely accepted for a long time. We believe this principle is fair to creators, necessary for innovators, and crucial for American competitiveness."
After learning of Balaji's passing, an OpenAI spokesperson said in a statement: "We are deeply saddened to learn of this incredibly tragic news today, and our hearts go out to Suchir's loved ones during this difficult time."
Currently, the reason for Balaji's suicide remains a mystery, but one point that has attracted media attention is that a legal document dated November 18 showed that Balaji was designated as a "custodian" or the person holding documents related to The New York Times' lawsuit against OpenAI. Balaji's sudden death has also sparked some conspiracy theories on social media.
