Column by DTU President Anders Bjarklev. Published in Altinget on April 22.
Denmark is facing a most serious security policy situation, which calls for immediate action. This is one of the first things that a new Danish government will need to focus on.
In this situation, the universities must contribute new knowledge and innovation in the field of defence and security, which is why the second government of Mette Frederiksen (known as the SVM government) in November 2025 allocated a historical amount of DKK 6.9 billion to research into critical technologies, security, and defence over the next four years. We appreciate the trust shown in us by the government.
However, the idea is for some of the funding for this research to be taken from the administration that makes it all happen: our research administration. I strongly urge the next Danish government to think twice about this.
Towards 2030, the universities are required to save a minimum of DKK 500 million on valuable employees who spend their working days advising, setting up, budgeting, planning, obtaining approvals, checking rules, communicating, coordinating, negotiating and recruiting. In other words: Employees who ensure the correct and responsible management of our research funding, leaving the researchers to focus on the core task.
For our researchers, these employees are indispensable. But for Boston Consulting Group, hired by the Ministry of Higher Education and Science to analyse possible savings in the public sector, they can easily be dispensed with.
Documents accessed by the media Ingeniøren shows that Boston Consulting Group sees artificial intelligence as a magical solution.
Less time, less quality
Specifically, the consulting firm believes that AI solutions will be able to free up 30,000 FTEs in the Danish public sector by 2035, including 3,900 FTEs in education and research, according to Ingeniøren.
Exactly how Boston Consulting Group has arrived at these figures, and with what consequences is not explained in any great detail. So, based on my knowledge of AI and the university world, let me present my analysis of the consequences:
Firstly, the implementation of new AI systems does not happen overnight. At DTU, we share the enthusiasm for new technology, but we also know that it takes training, evaluation, and supervision by real people for AI to work as intended. As we all know, a hammer doesn't drive a nail in by itself—and certainly won't hit the mark without a hand to guide it. The same goes for AI.
Secondly, the universities employ administrative staff for a reason. Their work underpins our research processes and ensures that high-quality new knowledge can be disseminated quickly and for the benefit of society. By discarding their many years of experience and organizational understanding and leaving it to AI, our researchers will be required to do everything that technology cannot handle. The impact of this is felt at UCPH right now.
In an article published in Science Report researchers explain how a comprehensive administrative reform at the university has resulted in less time for research and reduced job satisfaction.
This leads me to the third objection: With the proposed savings of DKK 500 million, it is hard to see how high-quality research can be maintained. Which takes us back to the beginning:
How can the universities continue to play a key research role in Danish defence and security policy after losing much the workforce involved in managing and supporting that very research? Everything is connected.
Learn from history
For us to get back on track and have constructive solutions pave the way for effective and well-thought-out improvements, a new government needs to change course: Drop the proposed savings plan and allow time for a thorough analysis of how AI can be implemented by the universities in a responsible and realistic way. The overarching goal must be to strengthen and support our research quality—not undermine it.
Yes, AI is a groundbreaking technology with huge potential. But it's not a magic solution that can free up hundreds of millions worth of resources just like that.
Ten years ago, the switch to digital post by the public sector in Denmark was also expected to yield huge efficiency and cost savings of more than DKK 1 billion. Instead, high operating and maintenance costs meant that the Treasury missed out on almost a quarter of a billion.
During the same period, the Health Platform was introduced in Capital Region of Denmark and Region Zealand. In practice, the magical platform, which should have saved the health service millions, was a huge disappointment and developed into a scandal costing the taxpayers hundreds of millions of kroner.
And let's not forget the tax IT systems scandal which—according to a National Audit Office report—has resulted in the Danish state now having to give up collecting outstanding tax of a staggering DKK 114 billion?
A new government should learn from these mistakes so that this time research does not end up paying the price.
DTU would like to support AI as a useful tool for us in our working lives. But we do not accept the rash assumption that AI can replace much of the workforce. Especially not at a time when responsible and critical thinking is one of Denmark's most important resources.
