AI is coming for your work, expert warns university staff

With management consultants predicting that up to one-third of work done today will be automated in the next five years – and universities under pressure to cut costs and do more with less – artificial intelligence offers a cheaper and more efficient way to keep higher education institutions running smoothly, claims an international higher education strategy expert.

Instead of trying to fight to protect traditional roles and jobs, Dr Ant Bagshaw, deputy chief executive of the Australian Public Policy Institute in Canberra, Australia, urges universities to embrace the unstoppable march of generative AI and accept that it is “more harmful to keep people in jobs that could be done better by robots”.

Bagshaw, who has an extensive portfolio of higher education strategy and consultancy roles in the United Kingdom and Australia, was speaking at a webinar on 10 November 2025 organised by the United Kingdom’s Higher Education Policy Institute (HEPI) following the publication of a new essay collection exploring how AI is reshaping the academic landscape.

The report, titled AI and the Future of Universities, was sponsored by HEPI and the University of Southampton and co-edited by Dr Giles Carden and Josh Freeman.

Productivity gains

Bagshaw said since the report's publication and his chapter asking whether it is time for university professional services to embrace GenAI, he had been accused of being “callous” about people's jobs and livelihoods.

He defended himself by saying he understands those critical of AI for its bias, intellectual property infringement, error and hallucinations, but “we should not pretend the system we have now is perfect [and] if our choice is between two flawed systems, we should choose the cheaper one”.

With the increasing pressure to do more with less, universities cannot afford to ignore the huge and unique productivity gains offered by AI, Bagshaw argued.

“As a learning sector at the frontier of the knowledge creation ecosystem, we need to embrace the changes and help people through them.

“A university does not exist to serve its staff but to serve its mission of education, research and knowledge exchange, and civic engagement,” Bagshaw told what HEPI’s director Nick Hillman said was the biggest audience ever attracted to a HEPI webinar, showing the intense interest in how universities should manage and use AI in the future.

Bagshaw said his favourite task is asking one of the GenAI tools – which has ingested the last ten years of the organisation's policy positions and media releases – what a mission group's response should be to the latest government consultation document.

“It is pretty good, not least because the same issues in the sector recur with enough regularity that the group already had a position. Generally, I get an answer which is around 80% there,” said Bagshaw.

In his chapter in the new HEPI report, Bagshaw cited research by McKinsey, the international strategy and management consultants, which claimed that 30% of hours worked today could be automated by 2030.

Roles likely to see the biggest productivity gains, according to McKinsey's research, include professionals in STEM (science, technology, engineering and mathematics) and education and workforce training, closely followed by creatives and arts management, and business and legal professionals and managers.

“GenAI is coming for the work of everyone involved in ensuring that universities keep running,” said Bagshaw.

Job market being transformed

Carden, chief of staff and chief strategy officer at the University of Southampton and one of the report’s co-authors, told University World News: “Graduate unemployment is at its highest level since 2018, and there is published evidence AI is transforming the job market.

“We are now in an era of what is termed ‘Hybrid Collective Intelligence’, where human insight meets AI. Those trained in AI will get jobs; those that aren’t will struggle.”

He noted: “I believe, if we fail to embed AI fluency across curricula, a university’s value proposition will erode.

“If graduates can’t secure jobs, why incur substantial debt to attend university? Why should governments keep spending billions?

“We are at an inflection point. Institutions that invest now in an AI fluency strategy – making it ubiquitous and mandatory across the curriculum – will thrive.”

Beware dumbing down human intelligence

In her chapter in the new HEPI report, Professor Rose Luckin, professor of learner-centred design at the UCL Knowledge Lab and member of the Computing College of the Engineering and Physical Sciences Research Council (EPSRC), called for policy interventions to ensure that AI serves as a tool for enhancing – rather than replacing – human intelligence.

Universities and others developing AI tools may be “dumbing down human intelligence” if they fail to get their approach to developing the new technology right, said Luckin.

She advised her webinar audience to “ignore the hype coming out of Silicon Valley about AI making our lives easier and effortless”.

It won’t, said Luckin, if we want to get it right. She warned that the “mainstream commercialisation of AI into consumer products and services” is sending counterproductive messages to students.

Luckin said the current AI revolution represented a “perfect storm” with three factors converging: vast amounts of data, advanced machine learning algorithms, and unprecedented processing power.

“These developments challenge what it means to be intelligent.

“As we develop increasingly sophisticated AI tools, humans need to become significantly more intelligent, not less,” said Luckin, who urged universities to use AI to extend, and not diminish, strenuous mental effort.

Understanding cognition

“We need to develop really sophisticated thinking and learning … so our students can understand what knowledge is,” Luckin noted.

Despite its remarkable speed and accuracy in processing information, AI lacks the “reflective and contextual understanding that defines human intelligence” – from critical thinking, emotional intelligence and creativity to academic knowledge, claimed Luckin in her chapter.

So, while AI will transform teaching and learning beyond typically narrow measures, such as test scores and completion rates, universities should use AI to help understand the nuanced development of cognition.

AI can help “identify patterns in how students engage with feedback, how they approach problem-solving or how their motivation fluctuates throughout a learning journey” and “lay the foundation for students to develop a more sophisticated understanding of themselves as learners”, said Luckin.

She made four policy recommendations to support higher education institutions developing comprehensive data strategies in her chapter: identify specific educational challenges AI can help address; identify sophisticated thinking capabilities students should develop that AI can track; define clear ethical guidelines for student data use; and ensure the interoperability of data systems across departments and institutions.

Impact on professional services

In his chapter, Dr Sudheer Parwana, a partner with PwC and OpenAI UK Alliance leader, urged higher education leaders to look beyond the transformative potential of AI for teaching, research and assessment and pointed to the impact AI can have on professional services, which he called “the administrative and operational backbone of a university”.

“From admissions to finance and from student support to human resources, these services shape the daily experience of students and staff alike.

“Yet they are often hampered by legacy systems, overstretched teams and manual processes. AI offers a chance to address these challenges, not by replacing human staff but by rethinking how they work,” wrote Parwana.

He said this was already happening, with the University of Glasgow using robotic process automation to tackle high-volume and repetitive administrative tasks, including automating document processing and data transfers across systems, with the result being “fewer errors, faster turnaround times and the liberation of staff to focus on student engagement”.

“Staffordshire University has introduced a digital assistant that lets students access their timetable, request documents, interact with student societies and even check in on their mental wellbeing.

“The chatbot operates 24/7, giving students immediate answers while reducing the burden on frontline staff.

“Rather than replacing human advisers, AI serves as a triage system, freeing staff to focus on complex or sensitive cases that require empathy and discretion,” said Parwana.

At Nottingham Trent University, a dashboard system tracks engagement metrics and enables early interventions, with the tools helping to support teams to act proactively, reaching out to students before problems escalate.

“They also allow institutions to allocate support more efficiently, focusing attention where it is most needed,” said Parwana, who accepted that the path to AI-enabled professional services is not without hurdles.

Chief among them is the challenge of AI maturity, with many institutions still in the early stages of adopting AI tools and staff lacking the skills or confidence to use them effectively.

Parwana pointed to a recent international survey which showed that only 37% of higher education institutions offer AI-related training for staff and only 1% have hired new AI talent.

“While upskilling is part of the solution, institutions also need to engage professional services staff in the design and implementation of AI tools. Without their insight, universities risk introducing systems that fail to reflect real workflows or reinforce poor processes,” said Parwana.

Nic Mitchell is a UK-based freelance journalist and PR consultant specialising in European and international higher education. He blogs at www.delacourcommunications.com.