The global AI governance landscape is evolving beyond traditional control versus empowerment debates, with experts emphasizing that responsibility must be allocated to institutions and people with knowledge, capability, and ability to prevent harm. According to recent analysis, the central challenge is not choosing between human control and individual empowerment, but ensuring that responsibility grows alongside capability rather than being transferred to those least able to exercise it. This shift comes as AI systems are developing capabilities that could fundamentally reshape human development, with potential impacts comparable to the steam engine, railways, electricity, and motorcars of the 19th century. The technology has already developed the capacity for coding and connecting directly with computers and the internet, moving toward artificial general intelligence (AGI) that could automate approximately 90% of current white-collar work.
AI systems are providing substitution for skilled workers in data processing while maintaining limitations in physical labor tasks, with the technology expected to create growing equalization of salaries for skilled and semi-skilled or unskilled workers. More concerning is the shift in income from workers to owner-suppliers of AI models, as returns come from data and information available online. This has raised concerns about fairer sharing of returns with information providers and users, representing what experts describe as the most important requirement for AI policy design. Recent developments show that 50% of American adults are more concerned than excited about the increasing use of AI in daily life, compared with only 10% who are more excited than concerned, according to Pew Research Center surveys. The legitimacy deficit surrounding AI governance is manifesting in public settings, with younger people confronting influential leaders with anxiety about AI's effects on their work and lives.
AI development is currently dominated by two countries - the United States and China - with about five widely used AI models each. Chinese models are cheaper to access and reaching approximately one billion direct AI users, with these models under government control. The political bias in AI responses reflects this orientation, with Chinese AI models providing China-biased replies when questioned in Chinese versus balanced responses in English. This creates concerns about AI becoming a tool for political influence and information control. Recent developments show that Anthropic researcher Jacob Coxon warned that AI could kill us all by the end of the decade, stating that neither OpenAI nor Anthropic is acting responsibly and that they are racing straight to self-improving superintelligence. OpenAI's Sam Altman and Elon Musk have since backed calls for slowing down AI development, acknowledging the harm that AI without proper guardrails could unleash.
Despite generating 20% of global data, India currently holds only a 3% share of global data. The country has approximately 150 data centres in Mumbai, Chennai, Bengaluru, Hyderabad and Noida, ranking 14th globally in the country-wise data centre hierarchy. India's data centre policy has announced a 21-year holiday for foreign firms providing cloud services based on domestic data centres, but promoting domestic AI models requires domestically-owned facilities with rigorous standards for water and energy use. Recent analysis suggests that a global AI safety-infrastructure fund, along the lines of the Global Climate Fund, should support evaluation, regulatory expertise and public-interest research in countries unable to participate effectively in AI governance.
India's AI policy must address four critical requirements: promoting global consensus for AI control to prevent undermining humanity, minimizing adverse social impacts on employment and income distribution, preventing dependency on foreign AI models, and developing domestically-owned data centres. The policy framework must recognize implicit politicization of AI output while promoting domestic AI development relatively free of ownership and government control. According to recent analysis, AI governance should be designed so that responsibility is visible and enforceable, with citizens retaining agency but not expected to manage risks they cannot reasonably understand or control. The ultimate test remains whether Indian nationals use sovereign AI models over Chinese and US alternatives, with responsibility existing at multiple levels - developers bearing design choices, deployers responsible for specific applications, regulators ensuring accountability mechanisms remain effective, and independent scientific institutions testing and monitoring AI systems.
