Once, a university degree was widely seen as a “ticket” to securing high-paying jobs and social mobility.
Now, as artificial intelligence (AI) promises to revolutionise the labour market, it’s university students and recent graduates who face some of the greatest uncertainty.
How do you pick a major or a career when it isn’t obvious what jobs will even exist in 10 years’ time?
Back in May, the chief executive of the AI company Anthropic, Dario Amodei, claimed AI could eradicate half of all entry-level white collar jobs over the next five years.
At this stage, it’s still up for debate whether AI will lead to such a mass wipe-out of graduate roles, or just change what these jobs look like.
Either way, we have a collective duty to prepare young people for an AI-driven world. Students, educators, employers and the government all have a role to play.
First foot on the ladder
In many white collar or “knowledge work” careers, the “lower rungs” of the career ladder have traditionally consisted of entry-level roles that centre on tasks such as data entry, routine report writing, document review or basic analysis.
These jobs were not only a rite of passage, but also a critical training ground for developing industry-specific skills, professional judgement and workplace confidence.
Many of these tasks are now at risk of being disrupted by generative AI.

This article is part of The Conversation’s series on jobs in the age of AI. Leading experts examine what AI means for workers at different career stages, how AI is reshaping our economy – and what you can do to prepare.
In the United States, the unemployment rate for recent college graduates is now higher than the broader unemployment rate.
Experts say this is due to economic uncertainty, high competition for jobs, and the slowing of white-collar job growth. But some argue the impacts of AI are also a factor, especially in fields like IT.
The International Labour Organization has published a list of “exposure indices” ranking a range of occupations from those deemed “not exposed” to generative AI to those that are highly exposed.
You can search yours below:
Replaced – or enhanced?
To help unpack some of AI’s impacts, it is helpful to quickly differentiate between the idea of “automation” AI, where jobs are replaced, and “augmentation” AI, which improves the output of existing workers.
Findings from a recent study suggest different kinds of work may differ in their exposure to these kinds of disruptions.
The study found in low-skilled occupations, automation AI could negatively impact new work, employment and wages. In high-skilled occupations, augmentation AI may raise wages and help create new work.
The study’s author, David Marguerit, suggests this could have negative implications for wage inequality.
Not the first time
AI is not the first technology to threaten the automation of young workers and early career tasks. From the introduction of calculators and computers to email and communication systems, technological innovations have steadily reshaped the nature of working roles.
Each of these innovations removed or transformed certain routine duties, often sparking fears of job losses, but also creating space for new responsibilities and skills. What makes the current wave of AI distinct is the breadth of cognitive and creative functions it can perform, and the speed at which these capabilities can be deployed across industries.
A 2022 study explored the potential risks of job automation for young Australians as they entered the workforce between 2009 and 2019.
Interestingly, its findings suggest young Australians often began in jobs at high risk of automation but reduced this risk by gaining qualifications, changing roles, or avoiding part-time or casual work.
Fewer options existed for avoiding jobs at high risk of change, such as data entry. Successful strategies for doing so were influenced by parental wealth, computer access, and ability to apply knowledge in new contexts.
This repositions the AI debate. Rather than predicting which jobs will last, we should tackle socio-economic divides by ensuring equal access to technology at home and in education, promoting the developmental use of AI and fostering critical reflection. For example, we could do this through structured classroom discussions about AI’s ethical and social impacts.
We also need to build a labour market that protects entry-level workers from soon-to-be automated roles to augmentation AI roles. In other words, getting them on the ladder.
What we all can do
How can we prepare for an AI-driven future? For those new to choosing career pathways, it’s worth looking at which industries are growing and which skills are hard to automate.
Jobs that require empathy, creativity and complex critical thinking are at less risk of AI automation, such as health care, education, creative arts, renewable energy and construction.
Policymakers and educators can also enhance the value of on-the-job training, such as internships and industry-linked projects, which have been shown to bridge theory and practice and improve career learning.
Recent research showcases a critical gap between the support students expect during placements and what they actually receive from workplace supervisors.
This means an investment in targeted upskilling, relevant AI-focused internships, AI-informed learning and teaching, as well as prioritising career learning as a core graduate outcome.
Authors: Rachael Hains-Wesson, Professor of Education and Associate Dean Learning and Teaching, RMIT University