Global artificial intelligence giant OpenAI has pushed back against copyright infringement allegations in India, asserting that its ChatGPT language model neither stores data nor undergoes training within the country.
The company made this statement during a hearing in the Delhi High Court on Wednesday, in response to a lawsuit filed by news agency Asian News International (ANI).
ANI has accused OpenAI of unauthorized use of its copyrighted content for training ChatGPT, alleging that the AI firm is reproducing its material without consent for commercial purposes.
Empanelment for Speakers, Trainers, and Cyber Security Experts Opens at Future Crime Research Foundation
OpenAI’s legal counsel Amit Sibal firmly countered ANI’s claims, arguing that India’s Copyright Act, 1957, does not apply to actions outside the country.
He further clarified that even if the Copyright Act were applicable outside India, OpenAI’s use of copyrighted material remains “non-expressive.” This means that any data used during AI training does not directly reproduce or display the original content to human users.
ANI has also raised concerns over ChatGPT generating inaccurate responses and falsely attributing statements, including fabricated ones, to the news agency.
These AI-generated errors, known as “hallucinations,” could potentially damage ANI’s credibility and contribute to the spread of misinformation, the agency claimed.
Now Open: Pan-India Registration for Fraud Investigators!
Denying ANI’s accusations, OpenAI informed the court that it had already blocked ANI’s website from being used for training ChatGPT since October 2023. It also emphasized that it has no servers, offices, or operational base in India, reinforcing its position that no copyright violation can be enforced against it under Indian law.
As part of its lawsuit, ANI is demanding Rs 2 crore in damages and a court order to prevent OpenAI from using, storing, or reproducing its content in any form.