Meity Seeking Concepts for Creating a Reliable AI Ecosystem and Tools

Nitin Konde, Varun Bhardwaj

Dec 17, 2024 — 3 min read

MeitY Invites Ideas for Building a Reliable AI Ecosystem

Startups and other pertinent parties have been asked to submit ideas for developing AI-powered tools that can identify deepfakes to the Ministry of Electronics and Information Technology (MeitY). The Ministry has requested five suggestions for the creation of domestic technical AI tools, rules, frameworks, and standards in order to establish a reliable AI ecosystem, in accordance with the expression of interest (EoI) issued by MeitY. As artificial intelligence (AI) advances make deepfakes more realistic, it is critical to provide deepfake detection techniques that protect society from possible disinformation and manipulation, as per the document released by MeitY. According to MeitY, in order to assess the authenticity of media and identify manipulation, the possible deepfake detection system should use advanced deep learning algorithms, provenance-based approaches, and other detection methods.

Additionally, it stated that in order to facilitate automated cross-modal content verification, offer real-time detection, and improve digital ecosystem security, the submitted deepfake detection technologies should be constructed in a way that makes them simple to incorporate into web browsers and social media platforms.

Laying Emphasis on Tool that can Differentiate Between AI Generated and Non AI Generated Content

The creation of a tool to distinguish AI-generated content from non-AI-generated content is another project the ministry hopes to carry out under the IndiaAI Mission. The EoI emphasised that in order to guarantee traceability and security and stop the creation of harmful and unlawful information, such a platform should contain distinct and undetectable identifiers in AI-generated content. The third initiative entails creating a platform that can assess how resilient AI systems are for high-stress situations like cyberattacks, natural catastrophes, and operational and data failures. Stress testing can guarantee readiness for high-stakes scenarios and offer useful insights into system flaws. According to the EoI, this theme focuses on the creation of instruments and techniques, like simulation-based testing or stress evaluation metrics, that are especially intended to strengthen the resilience of AI systems under pressure.

Setting Up of Ethical AI Framework

Furthermore, in order to guarantee that "AI systems respect fundamental human values, uphold fairness, transparency, and accountability, and avoid perpetuating biases or discrimination," MeitY has urged stakeholders to submit recommendations for "ethical AI frameworks" that provide an organised method. The report also stated that in order to facilitate the widespread use of AI, "it is necessary to create thorough AI Risk Assessment and Management tools and frameworks that are intended to detect, evaluate, and reduce cross-sectoral safety risks, guaranteeing that failures in one area are contained and managed without affecting interconnected sectors."

Applications are Invited From All Institutions

The EoI adds that entrepreneurs, academic institutions in India, research and development groups, commercial businesses, and startups are all welcome to apply for the projects. This comes days after S Krishnan, the IT secretary, allegedly stated that the Ministry was developing a system to assess the reliability and safety of AI technologies. Additionally, he stated at the time that the centre aims to emphasis innovation while avoiding restrictive rules. Earlier this year, Union IT Minister Jyotiraditya Scindia notably reiterated this opinion, stating that ethical considerations and a strong legal framework should govern the adoption of AI. The Indian generative AI (GenAI) ecosystem, which is home to 200 businesses and has raised over $1.2 billion in funding between 2020 and the third quarter (Q3) of 2024.