Opinion | As AI gets more powerful, humans and human values must remain in charge

As the franchise finale, Mission: Impossible – The Final Reckoning revisits past missions, showcasing how deadly weapons that threaten humanity have evolved – from viruses and nuclear weapons to AI, now cast not merely as a tool but the villain’s accomplice, even the puppet master. Meanwhile, Ghost in the Shell, the 1995 anime film enjoying a 30th anniversary rescreening, offers a more philosophical approach, exploring the blurred line between humans and AI through a cyborg’s existential quest, repeatedly asking: Who am I?

This year, tech giants including OpenAI, Anthropic and Google rolled out AI agents back-to-back, built to automate end-to-end workflows with minimal human input.

Heralding AI agents as “the next giant breakthrough”, OpenAI CEO Sam Altman has overseen two landmark releases this year. While Operator helps users fill out forms, order groceries and even complete purchases, Deep Research is designed to carry out complex, multi-step analyses at the level of a research analyst, completing in “tens of minutes” what would take a human “many hours”.

Meanwhile, from content recommendation and résumés screening to autonomous driving, AI is evolving from a silent assistant into the one behind the wheel. As more decision-making is delegated to algorithms – who gets hired, who might commit a crime, who deserves trust – this shift marks not just a technological leap, but a transfer of power away from humans.