The News Publisher-Tech Platform Relationship
News publishers rely on tech platforms like Meta for hosting content and generating traffic, which they monetise via advertising.
Tech platforms, such as Meta, act as intermediaries, driving traffic to news websites and taking a share of the advertising revenue.
The relationship has become contentious, especially as the field of Generative AI (GenAI) introduces new complexities in content usage and distribution.
Generative AI and Copyright Disputes
AI firms use vast datasets from the open Web, including news articles, to train models like OpenAI’s Large Language Model (LLM).
News publishers like The Atlantic license content to AI companies, while others, like The New York Times, take legal action over unauthorized use of copyrighted material.
In India, ANI has filed a lawsuit against OpenAI for using its copyrighted work without permission to train its LLM.
Generative AI and Copyright Disputes
AI firms use vast datasets from the open Web, including news articles, to train models like OpenAI’s Large Language Model (LLM).
News publishers like The Atlantic license content to AI companies, while others, like The New York Times, take legal action over unauthorized use of copyrighted material.
In India, ANI has filed a lawsuit against OpenAI for using its copyrighted work without permission to train its LLM.
The Legal Battle: ANI vs. OpenAI
ANI claims OpenAI used its content without authorization and argues that opting out doesn’t work due to third-party republishing.
OpenAI defends its use of content, stating that its models do not reproduce material verbatim, but rather modify it sufficiently under copyright exceptions.
ANI also highlights issues of fabricated responses by AI (e.g., false attribution of interviews), to which OpenAI has pledged to resolve.
ANI seeks an injunction to stop OpenAI from using its content and accessing it through subscribers
Implications
The dispute highlights challenges in balancing copyright infringement with fair use, especially in the context of AI model training.
In India, the concept of ‘fair use’ does not directly address AI training, creating legal uncertainty.
The absence of Text and Data Mining (TDM) provisions complicates matters for AI development.
The issue of territoriality in data storage raises concerns about data sovereignty, as AI platforms often store data across multiple countries, complicating legal accountability.
This lawsuit could set a legal precedent in India for determining the responsibility of AI developers in relation to the content used to train their models.
COMMENTS