On 21 November, Suchir Balaji celebrates his birthday. He’s found dead a few days later. Police calls it a suicide. His parents cry murder. Literally.

Two parallel investigations are on into his death. One by police in the U.S. and another by a private investigator hired by his parents.

Suchir Balaji was a young engineer. A U.S. national of Indian-origin, he had a promising career ahead of him. He worked with the best at OpenAI, the artificial intelligence research organisation best known for launching a chatbot called ChatGPT. In fact, it was Suchir Balaji who worked on developing WebGPT – a precursor to ChatGPT.

Today, ChatGPT is credited with accelerating the A.I. boom. OpenAI is valued at 157 billion dollars because of the success of ChatGPT. It’s pushed Microsoft’s valuation beyond 3 trillion dollars. Microsoft is a major investor and strategic partner of OpenAI.

What makes Suchir Balaji’s death controversial is that he quit OpenAI on a sour note. Just three months before his death. He accused OpenAI of plagiarism and copyright infringement.

Simply put, OpenAI was, allegedly, stealing content available online. All for free. Without paying any money to its rightful owner. Or without prior permission.

Companies such as The New York Times have since dragged OpenAI to court for violating copyrights. On 23 October, the N.Y.T. publishes an interview of Suchir Balaji. The article is headlined – “Former OpenAI researcher says the company broke copyright law”. It quotes him as saying – “If you believe what I believe, you have to just leave.”

Suchir Balaji publishes a blog (https://suchir.net/fair_use.html) on his personal website the same day as the N.Y.T. interview comes out. It is titled — “When does generative A.I. qualify for fair use?”. The second line of the blog reads – “… the process of training a generative model involves making copies of copyrighted data.”

A month later, N.Y.T.’s lawyer names Suchir Balaji in court-documents as someone who might have “unique and relevant” documents to support their lawsuit against OpenAI.

In a separate interview, Suchir Balaji tells the Associated Press news agency that he would try to testify in the copyright infringement cases. (Unlike the N.Y.T., the Associated Press entered into an arrangement with OpenAI in July 2023 to share access to select news content and technology.)

The A.P. interview quotes Suchir Balaji as saying — “It doesn’t feel right to be training on people’s data and then competing with them in the marketplace…. I don’t think you should be able to do that. I don’t think you are able to do that legally.”

OpenAI is being accused of doing something similar to what Napster did to music two decades ago. Napster was created by a 19-year-old student as a free service that allowed users to share songs. Its popularity threatened the profits of the recording industry and artistes, leading to lawsuits. Napster was sued out of existence but its impact can be felt even now in the music streaming industry.

Or what Google was, and still is, doing to news. Google uses news published by others to generate ad revenue worth billions of dollars for itself. Media companies have been complaining for years now that Google does not pay them fair money.

(The tide begins to turn, albeit slowly, after some countries and governments crack down on this practice. In 2019, the European Union passes a copyright law that requires Google and other online platforms to pay musicians, performers, authors, news publishers and journalists for using their work. Australia has since made such payments mandatory for tech companies such as Google and Meta which uses news generated by publishers to attract readers and advertising revenue. Canada introduces a similar legislation to ensure that online platforms pay fair remuneration for their content.)

The New York Times is not alone. Hollywood star Scarlett Johansson’s claim that ChatGPT used an imitation of her voice highlights tensions over rapidly evolving A.I. technology. A report published by The Guardian newspaper about the actor’s case says “people feel like their choice and autonomy is being taken from them”. The same report goes on to say that the Johansson row highlights a collapse of trust in A.I. OpenAI has since removed the said voice that had an uncanny resemblance to Johansson’s.

So, here’s the nub of the problem. OpenAI is, again, allegedly, making money off newspapers such as the N.Y.T. by using articles published by them to make ChatGPT more intelligent and efficient. OpenAI trains its Large Language Model using content it doesn’t pay for. That includes articles from the N.Y.T. and other publications and books.

Machine Learning and Large Language Model are a type of A.I. Machine Learning is defined as the ability of a machine to imitate or mimic human behaviour. Large Language Model uses Machine Learning to process and understand natural language. L.L.Ms are trained on large amounts of text data, such as books and articles, to learn how language works. L.L.Ms can perform a variety of tasks, such as, generating and translating text, and, answering questions. The allegation is that OpenAI scrapes or ingests millions of articles to make ChatGPT intelligent. ChatGPT then uses this information to produce text that not only rivals journalism companies such as the N.Y.T. but also mimics or spits out an article that readers would otherwise have to pay to read.

So, the question is — does OpenAI make a COPY of an already published article or generate an article INSPIRED by it? OpenAI argues that the legal doctrine of “fair use” protects its Training Model.

But concerns remain.

Like Suchir Balaji, Elon Musk has reservations about A.I. Musk tells a summit in October that A.I. could morph into something dangerous if it’s not developed carefully. Musk, who cut his ties with OpenAI in 2018, is suing it for switching to a for-profit model. OpenAI was founded in 2015 as a not-for-profit research organisation. Musk accuses OpenAI and Sam Altman of deceiving him by now opting for a for-profit structure. The “Godfather of A.I.” and 2024 Nobel Laureate Geoffrey Hinton has come out in support of Musk.

Musk has weighed in on Suchir Balaji’s death. “This doesn’t seem like a suicide,” he writes on X.

A message posted by Suchir Balaji’s mother on X says a private autopsy does not confirm the cause of death as stated by the police. She claims her son was killed and demands a proper investigation.