Leadership: Expectations concerning Artificial Intelligence challenge town halls

On paper it sounded like a simple task. As part of their daily services, four Danish municipalities delivered food to select citizens. Then they decided to join forces on a project that would introduce Artificial Intelligence (AI). This new technology would help analyse and optimise work processes in each of the municipalities, which in turn would help them reduce carbon emissions and save money.

Was it possible to shorten the routes? Or was a longer, but less busy route the better choice? Would it be possible to reduce the wear and tear of the vehicles? And was it possible to reduce the number of hours employees spent driving?

And did the individual municipality have the right combination of electrical, hybrid and petrol cars, and should they also introduce electrical bicycles?

Expectations that the intelligent robot (i.e. the algorithm – AI’s computer code) would come up with excellent answers were great. And as the project is still ongoing, we cannot assess its performance. Nonetheless, the two researchers who have followed the process are able to conclude that the implementation of such sophisticated technology is much more challenging than most people realise.

More work for the leadership

“Our task was not to assess whether the project itself was successful. Instead, we looked at the work carried out by the municipal leadership as they implemented this new AI technology,” Frank Meier, Postdoc, explains.

Meier conducts research on various digital aspects of leadership and has published a scientific article about the project in collaboration with his colleague Lise Justesen, Assistant Professor. Both are employed at the Department of Organization at Copenhagen Business School. One of their main conclusions is that the municipal leadership was faced with a task that for many reasons was greatly underestimated and significantly more complex than first anticipated.

There is a tendency to present AI as a relatively defined and complete solution, but that is not the case.

Frank Meier, Postdoc

Because even though politicians and everyone else talk about the many advantages of AI, according to the researchers from CBS, there is surprisingly little knowledge on what it takes to implement the technology. And in fact, it turned out that the leaders had their work cut out for them.

When AI and algorithms are to be introduced at town hall level, it brings many questions to the surface that no one is really prepared for. Or know the answers to.

Challenges are underestimated

“There is a tendency to present AI as a relatively defined and complete solution, but that is not the case. It turned out that there was much hidden work for the leaders to deal with and that the task required professional insights that the municipalities just do not have to a sufficient degree right now,” Frank Meier points out.

“First of all, there are all the technical questions – what must the municipalities specifically buy to fulfil their needs? This often turns into a tug-of-war between leaders who try to limit a project, while suppliers point to even more technologies that could be purchased,” Frank Meier elaborates.

And it is not only the new technology that requires deep knowledge. Municipalities are also navigating unchartered seas in other areas.

Is it, for example, possible to test an algorithm on data containing personal information? And would it be possible to deliver library books and food together – or are municipalities not allowed to share information on who borrows books and who has food delivered?

Digitalisation generates new tasks

“It involves more administration, technical equipment, leadership and legal work – for instance in terms of GPDR – than people realise. We can see that digitalisation generates new tasks,” Frank Meier emphasises and questions whether the municipalities are geared for this new reality:

“If solutions to societal problems are increasingly solved by AI, how will the executive and the city council handle these challenges? Their point of departure will still be classical themes such as eldercare, schools/youth, culture as well as roads and facilities,” he points out and then adds:

“Politicians may agree on reducing carbon emissions, but what do they know about organising and supporting AI and the extra work it carries? Currently, this task may be handled by the local IT department, but that is not the right place for it either.”

Chief executive calls for a greater focus

Pernille Halberg is a member of the The Danish Association of Local Government Executives and Chief Executive of Hørsholm Municipality. She is convinced that municipalities internally across Denmark must focus more closely on the leadership work required when implementing new technology.

Furthermore, in the municipalities, we basically lack sufficient knowledge about digitalisation.

Pernille Halberg, Chief Executive

“It concerns everyone from political leaders to general management, the heads of centre, and the leaders of people-centred units or town halls. Change management is difficult at the best of times, but changes in the shape of new technology challenge the process even more,” Pernille Halberg points out and then adds:

“Furthermore, in the municipalities, we basically lack sufficient knowledge about digitalisation. It is a new skill that we must acquire. In the future, we have to view digitalisation in the same way we view legal and economic skills – that is, as a prerequisite professional skill that will enable us to carry out our job as leaders.”

Metaphors oversimplify the challenges

Frank Meier points to the way politicians and others have created certain metaphors and narratives about digitalisation, which makes it difficult to talk about the disadvantages. For example, the narrative about the digital highway.

“The motorway is the frictionless mode of transportation we all dream about, rather than having to drive through the town centre at 2 PM on a Friday. It is seductive to talk about digitalisation like this, because it does not require us to understand the technical details. Or that we consider the many obstacles and bumps that we will encounter during the practical implementation.”

“Politicians have no problem using words like automatisation, optimisation and effectivisation. Words that generate expectations about the many issues that will be solved by digitalisation. And chief executives have to deal with this level of expectation,” says the CBS researcher, who has great respect for the work carried out by chief executives:

“However, the challenges they face in terms of AI is underestimated, and neither politicians nor AI consultants necessarily point out the red flags. Instead, as a society, we have created this narrative about how the implementation is a case of plug-and-play. It is this narrative we as researchers would like to problematise,” he states.

By Kent Kristensen, kk.slk@cbs.dk

---

The research article springs from the project Valuing Invisible Work, a project supported by the Independent Research Fund Denmark and anchored at the Department of Organization at Copenhagen Business School. It analyses the invisible work carried out in connection with the digitalisation of public organisations. Read more: https://www.cbs.dk/en/research/departments-and-centres/department-of-organization/weto-work-expertise-technology-and-organization/valuing-invisible-work-view