"My big concern is this 20-year transition period, where they’ll have winners and losers and a lot of disruption, and we have to get through that in order to get to that period of abundance," he said.
Gates, who chairs the Gates Foundation, said the pace of AI development over the past year has been remarkable, particularly in coding, biological research and the ability to perform increasingly complex tasks. However, he cautioned that governments and society were not adequately prepared for the resulting challenges.
"I’d say the response broadly hasn’t been adequate to the speed that the challenge is arising. I’m very concerned that the bad stuff will not be minimized and that a lot of good stuff could come fairly slowly if we don’t manage it the right way," he said.
AI could replace white-collar and blue-collar jobs
On employment, Gates said AI's growing capabilities could significantly reduce demand for human labour, initially in white-collar professions and eventually in jobs requiring physical work.
He pointed to accounting, telesales and customer support as areas where AI could perform tasks faster, more accurately and at substantially lower costs.
"As the AI achieves superhuman intelligence, the scarcity of white collar skills goes away," he said.
Gates expects humanoid robots to eventually extend this disruption to construction, agriculture, healthcare and other sectors dependent on physical labour.
He suggested that within about two decades, robots could become capable of building houses, growing food and performing surgeries, potentially reducing the need for human labour across much of the economy.
However, Gates acknowledged that the timeline remains uncertain and that technological advances could arrive sooner or later than anticipated.
He said a future of widespread abundance, where goods and services become significantly cheaper, would also raise difficult questions about how people find meaning and purpose when their economic contributions are no longer essential.
Cyberattacks and misuse pose immediate risks
Beyond employment, Gates identified cyberattacks and the misuse of AI in biological threats as major near-term concerns.
"We’ve crossed the threshold of capability that this is by far the most dangerous technology ever invented, and it’ll have pervasive effects in the economy and how war is fought," he said.
He expressed particular concern about AI models whose safeguards can be removed, potentially allowing malicious actors to exploit their capabilities.
Gates argued that leaving safety decisions entirely to technology companies would be inadequate, especially as competition among developers intensifies.
"Once you have multiple competitors, the level of responsibility can’t be with the companies. It’s completely obvious. The responsibility becomes the government’s responsibility," he said.
He called for mandatory oversight of advanced AI systems and stronger safeguards to prevent their misuse. He also suggested that highly capable models should operate on platforms that prevent users from removing essential security protections.
While supporting efforts to slow AI development if necessary, Gates said reaching a globally enforceable agreement would be extremely difficult, particularly because of international competition.
AI offers major opportunities in healthcare
Despite his concerns, Gates maintained that AI could deliver enormous benefits in healthcare, education and agriculture.
He said the Gates Foundation was already deploying AI in drug discovery, efforts to improve African-language capabilities, agricultural research and healthcare delivery.
"Healthcare discovery and delivery will be incredibly accelerated by AI, and that’s very exciting — better seeds, better livestock genetics, weather forecast for farmers in Africa, who never had any of those things," he said.
Gates also suggested that advances in AI-powered biological modelling could eventually help scientists understand drug side effects without relying on human trials, although he acknowledged that such capabilities do not yet exist.
However, he warned that the positive outcomes could take longer to materialise than some of the technology's immediate negative effects.
"Some of the things, like cyberattacks and the job market thing, even in the next two to three years, you’ll see significant negative effects in those two areas," he said.
Humanity must decide how to use AI
Gates also raised the possibility of reserving certain professions for humans, even when machines become capable of performing the same work more effectively.
He cited childcare, food preparation, art and sports as areas where societies might choose to preserve human participation.
He said the transition could require new policy choices, including shorter working weeks and early retirement, as governments seek to manage the changing relationship between employment, income and technology.
Rejecting the suggestion that he had become pessimistic about AI's future, Gates said he remained optimistic about its potential while recognising the scale of the disruption ahead.
"No, I’m not a doomer. I’m a strong voice to say that this is the most profound technology, both in its ability to do good and the rate of change that society will go through," he said.
Gates stressed that the challenge was not simply to accelerate AI development but to ensure that governments, researchers, industry leaders and society work together to minimise its risks.
"We have choices as humans to make about how we shape it," he said.