By The government is set to make changes to the Information Technology (IT) Rules of 2021, aiming to introduce regulations for both artificial intelligence (AI) companies and generative AI models.
Reports suggest that the amendments will require platforms using AI algorithms or language models to ensure they are free from any sort of bias. Sources familiar with the matter cited by ET mentioned that these changes will mandate that AI models or language algorithms trained on biased datasets or carrying inherent biases will be barred from public use.
Expected soon from the Ministry of Electronics and Information Technology, these amendments will not only focus on preventing bias in AI and language models but also potentially include regulations concerning deepfake and synthetic content. Furthermore, the rules might provide explicit guidelines for loan apps.
These amendments aim to expand the scope of trust and safety on the internet by ensuring adherence to specific parameters, such as caste, religion, community, and national security. Government officials have highlighted plans to engage in consultations to identify and mitigate biases beyond the recognized parameters, thereby enhancing safety measures.
Officials have also emphasized the need for AI models to undergo rigorous testing, including sandbox and stress testing, to ensure the absence of biases before they are introduced to the market.
The Digital India Act, intended to replace the IT Act of 2000, is not expected to be finalized before the upcoming general elections in 2024, according to Minister of State for Electronics and IT, Rajeev Chandrasekhar. This bill aims to create a secure and accessible internet for Indian users while also challenging the dominance of major tech corporations. Unlike the existing IT Act that provides safe harbor provisions for intermediaries, the Digital India Bill places responsibility on intermediaries for the content hosted on their platforms.
Earlier released presentations from MeitY underscored the pressing need for a specialized adjudicatory mechanism for online civil and criminal offenses. Despite earlier reports indicating a potential first draft release by the end of July 2023, the Bill’s progress seems to have slowed.
Among its provisions, the Bill is anticipated to introduce substantial penalties of up to INR 500 Cr for breaches. Additionally, the Bill is expected to define offenses like doxing, cybersquatting, astroturfing, and other online behaviors with clarity.
