Human-centred control of AI is both possible and necessary

The North Atlantic Treaty Organisation (NATO) encompasses, among other things, the world’s leading military and digital governance. As artificial intelligence (AI) becomes an ever more integral part of NATO’s infrastructure, what role is left for “the human in the loop”? 

According to a recent book chapter authored by Joanna Bryson, Professor of Ethics and Technology at the Hertie School, meaningful human oversight of AI systems is not only possible but necessary. The chapter appeared in the ebook Clicking the Pause: The Role of Transatlantic Cooperation in Artificial Intelligence Supervision, published in the series NATO Science for Peace and Security Series, E: Human and Societal Dynamics. 

In the chapter, Bryson reviews the AI regulatory strengths and strategies of the US military and the European Union and discusses recent apparent failures and other causes for concern regarding human-centred AI control. She concludes that the governance of AI in international security must adhere to ethical rules and that humans must be the ones to create and safeguard these rules. 

The publication contributes to ongoing international discussions on AI regulation, transparency, democratic accountability, and the role of transatlantic cooperation in shaping the future of responsible AI governance.