AI and robotics could replace as many as 40 percent of US jobs, McKinsey report suggests

A new analysis from the McKinsey Global Institute suggests that about forty percent of jobs in the United States could be carried out by artificial intelligence (AI) and robots if organisations chose to redesign how work is done. The report claims that existing technology is already capable of automating more than half of all work hours across a wide range of roles, spanning both manual and cognitive tasks.

The research highlights positions involving drafting, information processing and routine reasoning as among the most exposed. Hiring has already slowed in several of these areas, including paralegal work, administrative functions, office support and some programming roles. McKinsey argues that these tasks align closely with the strengths of current AI systems, which can generate, analyse and organise information at scale.

Physical and hazardous jobs are also considered vulnerable. Work in warehouses or involving machine operation is moving further within reach of robotics, which continue to improve in capability and reliability. By contrast, the report identifies a substantial share of roles that remain resistant to automation. Around a third of US jobs are said to depend on human qualities such as empathy, dexterity and judgement. In healthcare, for example, a large majority of tasks carried out by carers and nursing staff still require physical presence and an ability to respond sensitively to individual needs. Maintenance and repair work, which often takes place in unpredictable environments, is also thought to be difficult to hand over to machines.

McKinsey suggests that broader policy choices, as well as investment in implementation, remain the main barriers to widespread adoption. Its report, Agents, Robots and Us, argues that redesigning workflows rather than attempting to automate isolated tasks could unlock as much as 2.9 trillion dollars in economic value annually by the end of the decade.

The institute does not expect human skills to disappear from the workplace. Instead, it anticipates a shift towards people working in partnership with AI systems. Workers may spend less time preparing documents or processing routine information and more time interpreting outputs, making judgements and directing automated tools. Teaching and similar professions are expected to adopt hybrid approaches, using AI to handle administrative burdens or generate material while retaining responsibility for decision making and oversight.

The report also notes that new roles are emerging as organisations adopt more automation. These range from product managers who co-ordinate AI tools to specialists focused on safety and reliability. Despite concerns about job losses, the authors argue that artificial intelligence still depends heavily on human supervision, testing and support.