Font size:
Print
India’s AI Ambitions: Launching an Indigenous AI Model and AISI
Context:
India is set to make a significant leap in artificial intelligence (AI) with the announcement of an indigenous AI model and the establishment of an AI Safety Institute (AISI) under the Safe and Trusted Pillar of the IndiaAI Mission.
More on News
- Union Minister Ashwini Vaishnaw highlighted this initiative, reinforcing India’s commitment to AI governance and risk mitigation.
- In recent years, governments worldwide have realised the importance of addressing AI risks proactively, leading to the establishment of AI Safety Institutes (AISIs) in several nations.
Global Context and Need for AISIs
- Countries like the U.K., U.S., Singapore, and Japan have already set up such institutes to evaluate and mitigate the risks posed by AI, ensuring its safe and ethical use.
- U.K.: Unveiled the ‘Inspect’ open-source platform for evaluating AI models across multiple domains such as reasoning and autonomous capabilities.
- U.S.: Established an inter-departmental taskforce focusing on national security and public safety risks related to AI.
- Singapore: Focuses on content assurance, safe model design, and rigorous testing of AI technologies.
- These institutes underline the importance of technical rigour and international collaboration to ensure the safety and fairness of AI systems across borders.
India’s Approach to AI Safety and Indigenous Solutions
- India, with its unique socioeconomic landscape, linguistic diversity, and technological gaps, must prioritise AI solutions that cater specifically to local concerns such as AI inaccuracy and the risk of discrimination.
- India’s AISI is designed to operate on a hub-and-spoke model, collaborating with various academic institutions, startups, industry players, and government departments to address these challenges effectively.
- India-Specific AI Solutions:
-
- Data Representation: Indian startups like Karya are tackling the issue of unrepresentative data by empowering rural communities to create high-quality datasets in Indian languages.
- Multilingual AI: Other startups are advancing AI development in multiple languages, ensuring that AI technologies are inclusive and accessible.
- These local initiatives are essential for ensuring social equity and overcoming technical challenges unique to India.
AI Safety Initiatives and Indigenous R&D
- India’s AISI will focus on advancing indigenous research and development with a focus on Indian datasets. Some of the key initiatives under the IndiaAI Mission’s Safe and Trusted pillar include:
- Responsible AI Projects: Eight projects have already been selected, focusing on areas such as watermarking, ethical AI frameworks, risk assessment, and deepfake detection.
- Expression of Interest (EOI): A second round of EOI has been launched, addressing critical AI areas to ensure safe AI development and deployment.
- The AISI will play a pivotal role in fostering AI innovation while ensuring that these technologies adhere to the principles of safety, transparency, and inclusivity.
Global Alignment and Collaboration
- India’s AISI cannot operate in isolation. To ensure that its AI policies and frameworks align with international standards, India must collaborate with global AISIs and develop systems that facilitate global interoperability.
- Key Steps for Global Collaboration:
-
- Global Standardised AI Safety Taxonomy: A common, standardised taxonomy for AI safety is crucial to ensure consistent communication and assessment of AI systems across different countries.
- A global taxonomy would help stakeholders from multiple disciplines speak the same language when assessing AI systems and attributing responsibilities across the AI supply chain.
- International Notification Framework: India’s AISI should advocate for the creation of a global notification framework that allows AISIs to share information about the purpose and potential impact of AI models. This would enhance transparency and enable coordinated governance, ensuring that nations are prepared for the safe deployment of advanced AI systems.
- Leadership in the Global South: As a leader in the Global South, India has a unique opportunity to champion inclusive AI governance, especially for emerging economies that lack the resources to develop their own AISIs. India can play a leading role in co-developing AI safety frameworks and evaluation metrics for tackling local challenges in these regions.
- Global Standardised AI Safety Taxonomy: A common, standardised taxonomy for AI safety is crucial to ensure consistent communication and assessment of AI systems across different countries.
Supporting Frameworks and Collaborative Efforts
- India’s collaboration with UNESCO on AI readiness and the insights from the MeitY-UNESCO project will provide a strong foundation for the AISI. This partnership has already identified gaps in the ethical development and deployment of AI technologies, which will help shape the future AI safety frameworks.
- The IndiaAI Mission is already focused on critical themes such as machine unlearning, synthetic data generation, AI bias mitigation, and privacy-enhancing tools, all of which can serve as the building blocks for a robust AI safety ecosystem in India.
