The news platform was called “Training AI models with copyrighted data without consent” as copyrighted piracy, but the high-tech giants claimed they needed a huge amount of data
DNPA has pitched for the need for a government that ensures fair compensation for content producers and recognizes rights in a digital environment
Participants also discussed whether regulators should allow text and data mining. Option for content creators to opt out of frameworks
In an ongoing dogfight between publishers and the AI Giants, news portals and content platforms reportedly flagged the use of copyright data to train large-scale language models (LLMs) without permission.
According to the Hindustan Times, content industry representatives raised concerns during a two-day stakeholder consultation organized by the panel under the support of the Promotion of Industry and Internal Trade (DPIIT).
At the meeting, chaired by DPIIT Additional Secretary Himani Pande from June 19th to June 20th, the online news platform was called “training AI models on copyright data without consent” and copyright infringement. Meanwhile, the Tech giants argued that the model requires a huge amount of data. Many of them are copyrighted.
“The Digital News Publishers Association (DNPA) firmly believes that using digital news publishers' content without consent for subsequently generated AI applications such as AI training and subsequent search support and information purposes constitutes copyright infringement.
For context, members of the DNPA include the digital wings of India's largest media houses, including Hindus, Dainik Jagran, Network18, NDTV, Indian Express, Malayala Manorama, ETV, India Today Group, and Times Group.
DNPA representatives have sold the need for a regime to ensure fair compensation for content producers and recognize rights in a digital environment. Another participant reportedly flagged the creator as worried that he was “stripped for data.”
Meanwhile, the up-and-coming AI startup has urged the panel to secure a level play field. Citing their rationale, representatives said that large companies are already training AI models without facing the same level of regulatory scrutiny or limitations, but smaller players are still in the early stages of building LLM.
“Who is considered the actual owner of content generated by AI? If a user puts a lot of effort into creating a prompt leading to a specific output, will it make the user a legitimate owner, or does ownership lie to the creator of the AI model?” It reportedly asked a representative of an AI startup.
According to the report, participants also discussed whether the regulator should use the option to opt out of the framework to allow text and data mining (TDM). For beginners, TDM involves scraping large amounts of data, such as training AI systems, such as text and images from the Internet.
Among these, participants pitched the idea of a statutory licensing mechanism for AI training purposes. This could allow you to pay a “government set fee” and follow certain rules.
The development comes two months after it was reported that the centre had set up a nine-member panel to review existing copyright laws amid the rise in AI-related disputes. The committee will investigate legal and policy issues related to AI and their impact on Indian copyright law.
Following the stakeholder meeting, the panel will soon publish a working paper outlining whether India's copyright law needs to be updated to address the challenges raised by AI.
The fight over copyright
The issue came to the fore last year after Ani sued AI Juggernaut and ChatGpt Maker Openai at the Delhi High Court (HC). The press claimed that the AI chatbot “recreated verbatim or substantially similar excerpts” of the work in response to user prompts.
The petition also alleges that Openai misused ANI content for commercial profits by training LLMS using new agency content, adding that the use of content by AI majors could dilute market share.
In January this year, media outlets such as NDTV, Network18, Indian Express, and Hindustan Times also participated in the lawsuit against Openai. Under the DNPA's banner, the news publisher argued that the “action” of AI companies constituted “clear and current dangers to valuable copyrights” for DNPA members and other outlets.
In addition to this, the Federation of Publishers in India (FIP) moved Delhi HC earlier this year against Openai on behalf of all members including RUPA Publications, S Chand and Co, Bloomsbury, Penguin Random House, Cambridge University Press and more.
In February, music labels such as the T-Series, Salegarama, and Sony also expressed their willingness to take part in the ongoing copyright lawsuit against the ChatGPT developers of Delhi HC.
