Global Summit Calls for Secure and Trustworthy AI Development
Dozens of countries, including the United States and China, endorsed a declaration calling for “secure, trustworthy, and robust” artificial intelligence at the fourth annual AI Impact Summit.
The five-day gathering in New Delhi drew tens of thousands of participants, including leading technology executives, policymakers, and researchers. It marked the first time the global AI summit was hosted by a developing country.
86 Countries Sign Non-Binding Declaration
A total of 86 nations signed the summit declaration. The statement emphasized voluntary cooperation and industry-led initiatives rather than binding regulation.
It described generative AI as “an inflection point in the trajectory of technological evolution” and said advancing secure and trustworthy systems is foundational to maximizing societal and economic benefits.
However, the document stopped short of outlining concrete regulatory commitments.
Critics argued that the broad language and absence of enforceable safeguards limited its practical impact.
Amba Kak, co-executive director of the AI Now Institute, called the outcome “another round of generic voluntary promises,” arguing that the declaration appeared aligned with industry interests rather than public protection.
United States Shifts Position
The United States, home to companies such as OpenAI and Google, had declined to sign last year’s summit statement over concerns that heavy regulation could stifle innovation.
At this year’s summit, US delegation head Michael Kratsios reiterated opposition to sweeping global governance frameworks. Despite that stance, Washington ultimately joined the main declaration.
The US also signed a bilateral AI agreement with India focused on promoting entrepreneurship-friendly innovation.
Safety, Jobs, and Energy in Focus
The summit addressed both opportunities and risks tied to artificial intelligence.
Highlighted benefits included drug discovery, language translation tools, and productivity gains across sectors. At the same time, discussions centered on job displacement, misinformation, surveillance concerns, and AI’s growing energy demands.
The declaration acknowledged risks such as online abuse and potential misuse of advanced systems but framed responses around voluntary industry measures and appropriate policy frameworks.
On employment, signatories emphasized reskilling efforts to prepare workers for an AI-driven economy.
Energy efficiency also featured prominently. With AI data centers consuming vast amounts of electricity, the statement underscored the importance of developing energy-efficient systems.
Call for Binding Commitments
AI safety expert Stuart Russell described the commitments as limited but not meaningless. He argued that voluntary agreements could serve as a foundation for future binding legal protections.
He urged governments to build on the declaration to ensure AI development does not impose unacceptable risks on society.
Looking Ahead
The next global AI summit is scheduled for Geneva in 2027. Meanwhile, the United Nations is moving forward with an Independent International Scientific Panel on Artificial Intelligence, confirmed by the UN General Assembly with 40 members.
India used the summit to position itself as a rising AI hub, announcing ambitions to scale data center capacity and attract more than $200 billion in investment over the next two years. Major US technology firms unveiled new infrastructure and partnership agreements during the event.
The summit closed with a cautious consensus. Governments recognize AI’s transformative power. Agreement on how tightly to govern it remains unresolved.
Read more news, and follow us on Instagram
France’s President Emmanuel Macron and Brazil’s President Luiz Inacio Lula da Silva meet on the sidelines of the AI Impact Summit. Photo: AFP
