TL;DR

Former UK chancellor George Osborne, now heading OpenAI’s “for countries” programme, warned at the Delhi AI Impact Summit that nations failing to adopt AI risk becoming “weaker and poorer.” His comments drew pushback from African leaders and Mozilla’s head, who challenged the idea that only the US and China can build meaningful AI systems.

A US-or-China choice

Speaking at the fourth intergovernmental AI summit hosted by India’s Prime Minister Narendra Modi, Osborne framed the global AI race as a binary choice. Countries outside the US and China face “two contradictory feelings,” he said — fear of missing out on a technological revolution, and concern about national sovereignty when relying on AI systems controlled by foreign powers.

“Don’t be left behind,” Osborne told assembled leaders, warning that without AI adoption, workforces would become “less willing to stay put” as talent seeks opportunities elsewhere. His employer OpenAI is valued at $500bn.

The White House reinforced this framing. Senior AI adviser Sriram Krishnan told delegates: “We want to make sure the world uses our AI model,” while criticising the EU AI Act as hostile to entrepreneurs.

Pushback from the Global South

Not everyone accepted the two-superpower narrative. Mark Surman, head of Mozilla, called it “a false premise” that benefits US and Chinese companies. Kevin Degila, who leads AI and data for Benin’s digital agency, said his country was building its own systems that fuse American and Chinese technologies with local language datasets covering 64 languages.

“Anthropic and OpenAI don’t reach the farmers,” Degila said.

Rwanda’s ICT Minister Paula Ingabire said her government sought partnerships “that are going to be progressively less necessary,” avoiding lock-in to dependent relationships.

Looking forward

Former UK Prime Minister Rishi Sunak, now advising Anthropic and Microsoft, urged political leaders to treat AI as an “action this day” issue rather than a future concern. The summit continues with discussions on safety standards, which some experts worry are falling short as the most advanced AI systems develop amid White House resistance to regulation.