ARTIFICIAL INTELLIGENCE (AI) AND COPYRIGHT ISSUES

DIMENSIONS OF AI  DEVELOPMENT 

NITI Aayog's  roadmap on AI for Inclusive Societal Development in India proposes AI  driven initiatives to uplift India's 490 million informal workers by improving productivity, social security, market access, and skill development through inclusive digital infrastructure and multi stakeholder efforts.Artificial Intelligence AI  provides  computer program,the ability to think and learn on its own, and become a simulation of human intelligence. Generative AI facilitate both direct and indirect applications to work  on structured data. Commonly known direct applications comprise of leveraging natural  language for structured database generating different  types of data  including historical, tabular, time series data etc. Studies have shown that  AI  would be a force of good , but it can transfom into a most disruptive force as well. For the first time we are experiencing something that is smarter than  the smartest human growing at  a rapid pace of 54% per annum. Apart from Chat GPT,LLM and further developments are to take place in areas of Artificial General Intelligence AGI, Artificial Super Intelligence ASI and next Organised Intelligence OI.Chat GPT refers to Generative predetermined transformation ,an open AI chat box that understands and generates human like text for conversations ,answering questions,working content,coding and more learning from massive   data.Eventhough Artificial General Intelligence AGI and Artificial Super Intelligence ASI are yet to develop such future technological tools . ASI would possess key charectristics of superior cognitive ability,creativity and problem solving skills enabling it to self improve and potentially solve complex global challenges.On the contrary it also raises serious ethical issues about cultural,social and moral impact.

NEED FOR INTELLECTUAL PROPERTY PROTECTION 

During 2020s  rapid advancement of  deep learning based generativeAI models raised questions about copyright statusof AI generated works including copyright infringement.This include text to image models- stable diffusion and large language models including Chat GPT.In accordance with the ongoing global initiative to address legal complexities surrounding AI generated problems including AI generated claims of  authorship, originality, and ownership  etc.Popular deep learning models are trained on mass amounts of  media scraped from the net often utilising copyrighted material. Since legal jurisdictions mostly grant copyright to original works of authorship  by human authors the definition of originality is central to copyright status of AI. In this context the first working paper on AI prepared by Department  for Promotion of Industry and Internal Trade(DPIIT) India which has  ' Proposed a mandatory blanket license and Royalty system to regulate how generative AI systems using copyrighted material'. Addressing similar issues very actively the US  Copyright Office multi -part report on copyright right and Artificial Intelligence focus on part 1 addressing  digital replicas, part 2 copyright ability of outputs created using generative AI and part3.addressing generative AI training. Consequentlly whether in India, US  or elsewhere these developments highlights the  contemporary international debate on authorship, ownership and originality in AI generated works with policy makers ,researchers and scholars exploring various methods and models to balance both innovation and protection of human creativity. It may be noted  here that rapid strides made in areas like LLM in terms of quality, indepth information and sophisticated reasoning are facilitated by two factors a.continuosly improving Interactive advancement in applied machine learning .b.Growing access to text,data and multimedia used in training these models. AI firms are arguing that information on the internet should be freely usable for such training, even if reproduction and similar acts of syndication by any non AI entity would be subject to a license fee and consent by the content producer. This has led to fierce debate that pits the interests of AI hyperscalers and developers against content producers like- news,entertainment, book publishing etc. In this context only Department for Promotion of Industry and Internal Trade DPIIT's draft working  paper on AI and Copyright issues  proposals assumes importance .The proposal intend to resolve growing global tensions between online  content publishers and AI firms who train their LLMs with  a large quantity of  text data invariably scraped from the public internet.The  working paper suggested a "blanket licensing " framework where website data is crawled by large language models (LLMs) like Chat GPT would recieve royalty through a copy right society which could share money between them.

HOW TO PROTECT COPY RIGHT

Eventhough it is great significance to protect intellectual property of authors/creators, it may be detrimental to insist on consent for every use of their intellectual property. This make researchers and developers at the mercy of the IP right holder.Rapid advancement of deep learning based generative AI models raised questions about the copy right status of AI generated works and how copyright infringement occurs  while training or applying.  it includes different models- text  to image, stable diffusion, Large Language Models such as chat GPT. Most of the deep learning models are developed from mass amounts of media scraped from internet, mostly utilising copyrighted material. While assembling training data the sourcing of copyrighted works is likely to infringe the copyright holders exclusive right to control re production without exceptions given in relevant copyright laws, failing which it will treated  as copyright infringement.The definition of originality is significant because most of legal jurisdictions  grant copyrigh only to original works of authorship by human authors and the definition of originality is central to the copyright status of  AI generated works. 

In India AI  and copyright are a developing area.Our Copyright Act 1957 focuses on human authorship, which means  that purely AI made content is not protected unless significant human activity (like detailed prompting is involved with ownership shifting to the person created. Training AI  on copyrighted data without permission is potential infringement. But courts have not ruled definitely and Government is exploring licensing models. Core issues involved in defining authorship for AI assisted works and addressing data usage for training, with an expert panel reviewing potential reforms to balance both innovation and creators rights. As per Indian law for AI assisted works ,the human user(prompter /developer) can be the author provided their input shows "original creativity " or "creative control ".Using copyrighted material to train AI models without licenses leads to infringement. On the otherhand purely autonomous AI  output lacks copyright protection. However Government is considering a " hybrid model " to charge AI firms for using copyrighted data, distributing royalties to creators.It may be noted here that the compensation of copyright holders has been a contentious issue, with news publishers in particular suing AI firms around the world. 

CONCLUSION 

In brief solutions for an intellectual Property (IP) policy in AI, lies in  mandatory licensing framework that would provide AI data scrapping tools  to limitlessly scrape public information from the internet.At the same time it would ensure that non profit copyright society like bodies would collect payments from AI developers on the basis of revenue accrued to them through the  benefits of training their models on Indian  content producers data. Issues regarding how the royalty amount would be decided based on- small publishers who  put in effort and investment in their work, and recieving very large remuneration a piece when the big media house putting out hundreds of articles a day.Advantages of copyright in Al comprise of protecting AI's underlying code,algorithms, human  directed outputs,fostering innovation, by invecenticing  developers,enabling licensing for revenue and providing legal tools (like AI powered detection) to fight infringement. Such protections do offer exclusive rights, prevent unautorized copying and support creators by recognizing  their skilled input in guiding.As India's AI prospects are booming due  to huge talent pool,strong IT sector and potential AI driven growth of sectors like Pharmaceuticals, manufacturing, automobiles  health care,  finance and agriculture. The focus is on achieving inclusive growth  through indigenous models and ethical framework. In any case ultimately  copyrighted AI must result in optimum development of the economy in terms of jobs, revenues to the government and inclusive development .