A global declaration urging the development of secure AI drew support from 86 countries at the AI Impact Summit in New Delhi. The statement emphasised the need for secure AI and outlined voluntary initiatives without binding regulations. The five-day summit highlighted both the opportunities and risks of generative technologies.
Secure AI at the Center of Global Declaration
The declaration described generative AI as a turning point in technological evolution. It argued that secure AI is essential to building public trust and ensuring the fair distribution of benefits. It stressed that AI must evolve in ways that support economic growth while limiting risks such as misinformation, surveillance misuse and potential biological threats.
Voluntary Measures Over Binding Commitments
The document focused on non-binding cooperation including international research pooling and industry-led safety measures. It stopped short of commitments to regulate or enforce limits on AI development. Analysts noted that officials had already expected limited outcomes due to broad agendas and previous vague statements in France, South Korea and Britain.
US Participation Marks a Shift
The United States, which avoided endorsing last year’s declaration, signed the New Delhi statement. US delegation head Michael Kratsios stated that Washington rejects global governance of AI and prefers innovation-driven approaches. The US also signed a bilateral agreement with India aiming to support an AI environment friendly to entrepreneurship. India delayed the declaration’s release by one day to secure wider international support.
Concerns Over Jobs, Energy and Safety
AI’s growing impact on employment and energy infrastructure was a major point of discussion.
Key concerns included:
• Job disruption and the need to retrain workers for AI-driven economies
• High energy demand from large data centres and the need for efficient systems
The declaration encouraged countries to invest in reskilling programmes and highlighted the urgency of reducing AI’s environmental footprint.
Civil and Expert Reactions
Critics such as Amba Kak from the AI Now Institute said the declaration lacked meaningful commitments and largely reflected industry priorities. AI safety advocate Stuart Russell argued that voluntary steps are a starting point but that countries must pursue binding measures to reduce unacceptable risks.
India Positions Itself as AI Hub
India used the event to signal its ambition to compete with major AI powers. The government expects more than 200 billion dollars in AI-related investments over two years supported by expanding data centre capacity and new nuclear-powered energy infrastructure. US tech companies announced several major deals and projects during the summit. Despite some logistical issues at the vast venue, attendance remained high with many high-profile industry figures. The next global AI summit will take place in Geneva in 2027 while a new UN scientific panel begins work on governance.
Conclusion:
The New Delhi summit produced broad agreement on the need for secure AI though without binding commitments. Countries signalled interest in cooperation while debate continues over how to balance innovation with public protection.
