Delhi High Court just greenlit AI training on copyrighted content under India's fair dealing laws. This sets a crucial precedent, directly clashing with publishers worldwide who demand licensing fees for their data. Indian news agencies now face an uphill battle defining content value in the LLM era.
How We Got Here
ANI filed the copyright infringement suit earlier this year, alleging ChatGPT reproduced its news reports and caused reputational harm. This came amidst a global wave of similar lawsuits from major publishers against AI companies in the US and EU.
The Numbers
- Court held OpenAI's storage of ANI's works falls under Section 52(1)(a) Copyright Act, thus not infringement.
- Justice Amit Bansal observed ChatGPT's outputs were not "substantially similar" to ANI's, denying Section 51 infringement claim.
- ANI failed to prove "memorization regurgitation" of its content by ChatGPT.
- The ruling prioritised "irreparable injury" to OpenAI and "the public at large" if an interim injunction were granted.
- The court ruled in ANI's favour on territorial jurisdiction, meaning the case can proceed in India.
What Happens Next
🇮🇳 Why This Matters for India
For Indian digital news publishers in Delhi and Hyderabad, this ruling weakens their bargaining power in demanding licensing fees from large language model developers.
The Take
This interim ruling is a clear win for AI developers: it implies fair use provisions are broad enough for training data. Publishers expecting a quick payout from LLMs just got a reality check on Indian soil.
Source:
MediaNama ↗