This essay traces the evolution of AI safety institutes around the world, explores different national approaches, and examines the need for an AI safety institute in India.

In this podcast for "The Citizens Talkshow," Amlan Mohanty joins Anirudh Narayan and Mayur Bommai to discuss a wide range of topics related to AI, including its definition, impact on businesses and society, and the evolving landscape of AI policy both globally and in India.

Carnegie India recently organized a closed-door meeting in New Delhi to discuss India’s approach to compute in the context of its national artificial intelligence (AI) strategy. The meeting was attended by government officials, technology executives, startup founders, and academic researchers. Carnegie India’s experts also attended meetings organized by technologists based in Bengaluru on their efforts to promote open access to compute. Below is a summary and analysis of the key takeaways from these meetings.

This essay provides three reasons why the world needs a global AI compact, what it will hopefully achieve, and the role of different stakeholders in this process.

This essay discusses how India can address its “compute conundrum” to develop a compute strategy that aligns with its AI strategy.

This essay shares insights on key elements of India’s AI strategy and outlines some of the trade-offs involved in balancing risks with opportunities.

Given the importance of AI under the iCET framework, this essay outlines five areas where India and the United States could collaborate to advance their partnership on AI.

DPIs have the potential to transform how businesses deliver their services, given the benefits of scale, efficiency, and innovation. But leveraging the power of DPIs in a sustainable way will require effective collaboration between businesses and policymakers.

As different countries work toward adopting a global definition for DPI, it is important to implement a principle-based approach built on consensus until such a definition is adopted.