Artificial intelligence is commonly presented as either a technological miracle or an existential threat. Yet the most immediate question is political: who will own and control it, and for whose benefit?

AI is being developed by corporations, rival governments and military establishments competing for profit and strategic advantage. Meanwhile, automation is placing pressure on capitalism’s dependence on wage labour. If machines can produce more goods and services with fewer workers, society becomes richer in productive capacity while many people lose the wages they need to live. Technology that could liberate humanity from exhausting and repetitive work instead threatens unemployment, poverty and greater inequality.

The answer is not to halt technological development, which is neither realistic nor desirable, but to bring AI under democratic control. Frontier systems should undergo mandatory independent safety testing and be governed by enforceable international standards. Governments should also establish publicly funded, genuinely open AI research centres and a universal digital library, giving everyone access to humanity’s accumulated knowledge while fairly remunerating writers, researchers and artists.

Countries outside the United States–China rivalry, including Australia, could cooperate to create this public alternative. Together, they could use their influence over semiconductors, computing infrastructure and other critical technologies to enforce common standards.

AI is the product of generations of collective knowledge and labour. It should become a global public utility, used to shorten working hours, eliminate drudgery, expand access to knowledge and improve human life – not to enrich a handful of corporations, extend state surveillance or strengthen military power.