Table of Contents
Context
The IndiaAI Mission has selected eight Responsible AI Projects against the Expression of Interest (EoI) floated under the Safe and Trusted AI Pillar of the IndiaAI Mission.
About Project
- The IndiaAI Mission, under the Digital India Corporation and Ministry of Electronics and IT (MeitY), has selected eight Responsible AI Projects from over 2000 proposals submitted by academic institutions, start-ups, and research organisations.
- These projects aim to advance the responsible development, deployment and adoption of AI technologies in India by creating indigenous tools, frameworks and guidelines that reflect the country’s unique challenges and opportunities.
About selected projects.
- Machine Unlearning: Develop techniques to enable AI models to “forget” specific data points, enhancing privacy and security.
- Synthetic Data Generation: Address data bias issues by generating synthetic data to ensure AI fairness.
- AI Bias Mitigation: Focus on reducing bias in AI systems, particularly in sensitive fields like healthcare.
- Explainable AI: Promote AI transparency by making AI models more interpretable and ensuring data privacy, especially for security-related applications.
- Privacy-Enhancing Tools: Create robust privacy-preserving machine learning models for sensitive applications.
- Ethical Certification: Establish a toolset, “Nishpaksh,” to evaluate AI models for fairness and ethical behaviour.
- Algorithm Auditing: Build frameworks for auditing AI algorithms to assess and improve accountability and fairness.
- AI Governance: Develop a governance framework to assess and monitor the safety, transparency, and ethical risks of large AI models like LLMs.
IndiaAI Mission
|