Scientists and regulators are concerned that the breakneck pace of AI development could lead humanity to lose control over the technology, with unpredictable consequences. There is an urgent need to strengthen international coordination to prevent such risks.
AI offers enormous potential to promote sustainable development, improve learning and work efficiency, strengthen healthcare systems, and enhance climate change resilience. However, an increasing number of people directly involved in developing the technology have warned that the pace of AI development is far outstripping human understanding of its risks. AI development is entering a new phase, in which systems are no longer limited to answering questions or generating content on request, but are increasingly capable of planning independently, using tools, and carrying out sequences of actions with less human intervention. This transition is fundamentally changing the debate over AI safety and control. Amid growing concern among governments, businesses, and the technology sector over digital sovereignty, cybersecurity, and AI governance, GovTech 2026 in South Africa placed particular emphasis on AI control as the technology's rapid evolution raises serious questions about future risks, while highlighting the need for a governance model that can both protect the public interest and remain flexible enough to adapt to technological change.
Debates over AI are no longer confined to Silicon Valley but have spread across the globe. European leaders have issued warnings about AI with varying degrees of severity, but all have called for greater control over the technology. Spanish Prime Minister Pedro Sanchez recently warned of devastating consequences if humans lose control of AI. He presented a 12-month plan aimed at creating “a new social contract for AI” in Spain in cooperation with civil society. The agreement must provide people with security in the face of changes to work and daily life that the technology will bring. Under the UK's upcoming G20 presidency in 2027, UK Prime Minister Andy Burnham said the country would place AI at the centre of its agenda, striving to establish a unified set of principles and standards to enhance the transparency and accessibility needed for nations to better address AI-related challenges.
Many leading figures in the AI industry support greater cooperation, regulation, and a slowdown in the development of the technology. Yoshua Bengio, a Canadian computer scientist who received the 2018 Turing Award, and fellow AI pioneer Geoffrey Hinton have even said that humanity is gradually “losing control” of AI, while stressing the need for safeguards on a scale and with an ambition comparable to mechanisms for controlling nuclear weapons.
Amid the urgent need to establish tools for controlling AI, United Nations Secretary-General Antonio Guterres said international coordination is needed to ensure that AI operates safely, transparently, and responsibly. This is not merely an issue for technology companies or regulators, but also a matter for businesses, the research community, international organisations, and civil society.
