Why in the News?

India's first AI Data Bank to boost innovation by offering researchers, startups, and developers access to diverse datasets for scalable AI solutions was launched recently.

More on the News

• The Data Bank will strengthen national security by enabling real-time analysis of satellite, drone, and IoT data for AI-driven disaster management and cybersecurity. 
• Additionally, India's National AI Strategy promotes innovation, ethical governance, and global collaboration, with partnerships to advance AI in healthcare, agriculture, smart cities, and space exploration.
  • India has established itself as a global leader in technology-driven governance and Digital Public Infrastructure (DPI) in last few years.
    • Digital Public Infrastructure (DPI) drives digital transformation and improves public services, helping countries meet national goals and accelerate Sustainable Development Goals. 
• All these are crucial components for overall governance architecture and its linkage with AI in India. 

Potential of AI to transform governance in India 

• Efficient Service Delivery: AI automates public services, reducing government workload and improving service quality. For ex-
  • Education: AI can transform India’s education system by enabling personalized learning, creating smart content, and automating grading and assessments.
    • For ex- NCERT has listed a set of 31 metadata elements to be tagged to each resource available in its NROER (National Repository of Open Educational Resources) repository.
  • Healthcare: AI is transforming healthcare by improving delivery and accessibility, particularly through telemedicine in remote areas.
    •  For ex- NITI Aayog with Department of Bio- Technology (DBT) aims to build database of cancer related radiology and pathology images for effective use of AI in cancer management.
  • Agriculture: AI provides predictive insights for weather, pest management, and resource use, benefiting farmers. 
    • For ex- National Pest Surveillance System utilizes AI and Machine Learning to detect crop issues, enabling timely intervention for healthier crops.
• Inclusivity and Accessibility: AI-powered DPI systems bridge gaps, particularly in a multilingual country like India.
  • For ex: Bhashini platform uses AI to provide government services in regional languages.
• Data-Driven Policymaking: AI analyzes large datasets for evidence-based policymaking, improving transparency and policy effectiveness.
  • For ex: The India Urban Data Exchange uses AI to optimize urban services like traffic management and waste disposal. 
• Judicial Efficiency: AI improves judicial efficiency by automating case management, prioritizing cases, predicting outcomes, and streamlining legal research. 
  • For ex-SUVAS (Supreme Court Vidhik Anuvaad Software) is an AI-based translation tool that bridges language gaps in legal proceedings.
• Disaster Management: AI-driven systems like the RAHAT (Rapid Action for Humanitarian Assistance )app help predict natural disasters, such as floods, by providing early warnings and supporting evacuation, search, and rescue operations during emergencies.

Challenges in AI Integration for Governance

• Fragmented Data across government departments: For ex- The National Data Governance Framework Policy (NDGFP), which aims to standardize the management of non-personal and anonymized data for data-driven governance, has not yet been implemented.
• Infrastructure Gaps: Poor internet, storage, and computing in rural areas, along with a lack of robust cloud infrastructure for AI, create a digital divide.
  • For ex- as of 2023, 45% of India's population, still lacks internet access, according to a study by IAMAI (Internet and Mobile Association of India).
• Regulatory Frameworks: India lacks AI-specific laws, unlike the EU's AI Act, raising concerns about ethics, data privacy, and accountability.
• Skill Gaps: A demand-supply gap of 140,000 AI professionals in India, as reported by NASSCOM, limits growth.
• Data Privacy: AI’s reliance on sensitive data increases breach risks, as seen in the Aadhaar data leak on the dark web affecting 81.5 crore Indians.
• Weak IP Regime: India ranks 42nd in the 2024 IP Index, offering limited protection and incentives for AI innovation.
• Ethical Biases: AI systems trained on biased data can produce discriminatory outcomes, raising ethical concerns about fairness and accuracy.

Way Forward

• Risk Management and Ethical Oversight: AI must be dynamically assessed and monitored, with human oversight to prevent biases and manage risks.
• Data Sovereignty and Privacy: Ensure data privacy and compliance with the Digital Personal Data Protection Act, especially for cross-border data.
• Bias, Fairness, and Transparency: AI must be fair and transparent, with audits, fairness metrics, diverse datasets, and "model cards" for critical sectors.
• Education and Skill Development: Expand initiatives like INDIAai FutureSkills to provide AI education in underserved areas.
• Public-Private Collaboration: IndiaAI Compute Capacity aims to build a scalable AI ecosystem with 10,000+ GPUs to support AI startups and research.
• Cybersecurity and Monitoring: Use AI for real-time threat detection and continuously refine AI policies.

Participatory AI Development and Governance Participatory Approaches in AI Development and Governance Paper released by IIT-Madras. About Participative AI (PAI) It refers to the involvement of a wider range of stakeholders than just technology developers in the creation of AI systems.  Core tenets of PAI are derived from participatory governance. (see box) Need: Progress in AI and its deployment by public and private actors, like Facial Recognition Technology in Law enforcement, etc.  Benefits of PAI: Counter unilateral, top-down decision making; Mitigate risks like bias, discriminatory output, etc.; Feedback loops for flagging technical glitches and post deployment impact assessment; Enhance trustworthiness of AI Systems through minimal false positives and false negative. Challenges with PAI:  Co-optation: Domination by select dominant actors to serve their vested interests.  Limited participation of non-experts.  Participatory washing and tokenism: Stakeholder participation merely done for formal compliance. Transparency Paradox: Information shared about algorithms can be misused by malicious actors.