Runaway AI warning! Tech pioneer urges world to slow down now

Speaking at the Digital World Conference (DWC): AI for Social Development   – co-organized by the UN Research Institute for Social Development (UNRISD) – Professor Hinton stressed that rapid advances in AI must be guided more carefully to serve societies – rather than undermine them.

“If you ever went out with a car that had no brake, boy, you are in trouble if you go down a hill,” he told delegates. “But you're in even more trouble if there’s no steering wheel.”

His remarks came during a busy week for AI policymaking, as governments and UN panels stepped up discussions on governance, inclusion and risk management, amid the growing integration of artificial intelligence across the global economy and society.

The haves and the have-nots

The pace of AI’s growth is staggering. According to UN Trade and Development (UNCTAD)’s Technology and Innovation Report 2025, the global AI market is projected to grow from $189 billion in 2023 to $4.8 trillion by 2033, an economy larger than Japan’s, built in a single decade.

Yet the capacity to build and shape it remains in the hands of just a few economies and firms, UNCTAD Acting Secretary‑General Pedro Manuel Moreno warned at the Commission on Science and Technology for Development (CSTD), also meeting this week.

That concentration risks deepening global inequalities. Doreen Bogdan‑Martin, Secretary‑General of the UN International Telecommunication Union (ITU), pointed out that generative AI adoption in the industrialized ‘Global North’ is growing nearly twice as fast as in the developing ‘Global South’.

“Left unaddressed, this is a second great divergence - widening the gap between countries shaping artificial intelligence and those merely consuming it” - Ms. Bogdan‑Martin said, adding that gaps in infrastructure, investment and capacity cannot be closed by any single country or organization alone.

This week’s flurry of international activity on AI and digital technology in Geneva and beyond, reflects the international push to ensure that all countries can benefit and regulate Artificial Intelligence as it increasingly shapes our economies, societies and daily lives.

Distinct areas of discussion are becoming clear.

While the focus of the Commission on Science and Technology for Development is on global‑level digital policymaking, discussions at the AI For Social Development Conference underscored the need for transparent, accountable and rights‑based AI governance to address risks such as bias, opaque algorithms and having large volumes of data concentrated in the hands of just a privileged few massive corporations.

Participants at the World Conference – convened by UNRISD and international NGO, the World Digital Techology Academy – examined AI’s growing role in social protection, labour markets, education and the green energy transition, while stressing the importance of protecting vulnerable groups and ensuring the benefits of technological change are shared more fairly.

Data-driven approach

Any proposals for AI governance must be data-driven and this is the fundamental work of the UN’s Independent International Scientific Panel on AI, which convened its first in‑person meeting in Madrid on Wednesday.

Opening the Scientific Panel’s first in‑person meeting in Madrid, co-chair Maria Ressa explained the group’s mandate to provide an independent, scientific and authoritative assessment of how AI systems are shaping societies.

Ressa, a Nobel Peace Prize laureate and campaigning Philippines journalist, warned that increasingly powerful AI tools are accelerating the undermining of democratic systems using “narrative warfare” in which falsehoods are manufactured and amplified at scale; the weakening of institutions such as the media and courts; and, ultimately, strategic corruption once accountability erodes.

Its findings will inform the discussions of another key UN AI initiative – the UN’s Global Dialogue on Artificial Intelligence Governance – which meets in July, also in Geneva.

A worldwide discussion

The Scientific Panel’s findings which Ms. Ressa co-chairs with renowned Canadian computer scientist Yoshua Bengio will inform the discussions of another key UN AI initiative – the UN’s Global Dialogue on Artificial Intelligence Governance – which meets in July, also in Geneva.

The Global Dialogue brings together all 193 United Nations Member States, the private sector, civil society, academia and the tech world to share best practices and build common approaches to AI governance.

“The policy conversation will be science and evidence-based, pooled perspectives, scientific perspectives from a multidisciplinary lens from across the world,” says the UN Special Envoy for Digital and Emerging Technologies, Amandeep Gill.

“This is how policy discussions should be, and the UN is very proud to facilitate this first ever such confluence of science and policy in a fast-paced emerging technology.”