This month, Tech Monitor found itself back in the northern latitudes of Scandinavia, almost a year after it consulted the region’s IT leaders about the efficiency-creating potential of AI for their businesses. Our first stop was Copenhagen, to take stock of AI deployments throughout Denmark and to explore how companies in the country are balancing power, performance and place.
What became clear during a lively two-hour discussion in Copenhagen’s Nimb Hotel is that maturity levels have risen. Today, most organisations are either deploying AI applications or are in the advanced stages of moving from proof of concept to rollout.
The merits of generative AI, especially when it comes to serving internal users and bringing efficiencies to internal processes, is clear. Less apparent is what AI means for the future of the workforce, coding, or the IT department itself. As a result, these existential questions – rather than the day-to-day concerns about infrastructure, compute power and the merits of cloud computing as home for AI experimentation – dominated the roundtable.
The end of the IT department?
One of the senior technology leaders present offered a vision of the future organisation that doesn’t include a traditional, centralised IT function. Given how GenAI is democratising software engineering, he argued that there is an opportunity to federate application development to every organisational function that requires it, from marketing to human resources and from finance to operations.
Not only do these functions know best what they need (and control their own budgets), AI now allows them to build what they need, too. The net result? Devolved IT leaving the proponent of this argument not only talking himself out of a job but threatening the employment status of most people around the table, too.
Unsurprisingly, not everyone was ready to accept this future vision. As another attendee pointed out, multiple departments responsible for their own development means multiple processes, multiple visions of business need, and, most likely, multiple versions of the same software tools. Another attendee put it simply: “It’s a disaster for compliance, complexity, and cost.”
Not everyone thought this way, however. Some felt compliance could be centralised in this federated model and that the IT function could be reinvented as a quality assurance-led department.
The future of coding
If not the end of the IT department, what about the demise of the human coder? Most organisations represented around the table – drawn from sectors including financial services, healthcare, and retail – are currently augmenting human coders with coding co-pilots. Broadly speaking, they are automating the “grunt work,” leaving software engineers to specify needs, build requirements, and validate outcomes (though the co-pilots happen to be pretty good at validation, too).
That is the division of labour, for now at least. But what if we no longer need developers at all? That question was posed by an attendee who, like others around the table, has spent his career being taught the value of coding.
Despite this, he wasn’t sentimental about its potential demise. Rather, he questioned an ongoing necessity in its manual incarnation. Another attendee compared today with Johannes Gutenberg’s invention of the movable-type printing press in the mid-15th century. For the monks whose job it had been to transcribe holy manuscripts, new technology rendered their role – but, critically, not their overall purpose – redundant.
For the tech optimist, automation opens up new opportunities. Human expertise gets deployed elsewhere, more effectively. And, to extend the comparison, in at least one telling of the arrival of the Gutenberg Press, few 15th-century scribes lamented the passing of a tedious job that took place in damp, cold conditions and increased the likelihood of poor and deteriorating eyesight.
The talent gap
Beyond questions of potentially deskilling a future generation, some organisations are struggling to plug wider gaps in their AI talent pool. As one attendee put it, it’s no longer a problem for firms to figure out where AI can help. “We have 140 use cases,” they said. Rather, “it’s about figuring out how to do it”. For her, it is now a question of how to identify and prioritise talent need. Data modelling, validation and implementation are among the skills that are in short supply, she said. Legal and compliance oversight capabilities are also lacking.
Elsewhere, organisations are seeking to reskill their existing workforce to reflect changing requirements. One, for example, is committed to providing at least ten hours of training per employee per year. The mission: to transform the organisation so it is AI-ready. This is perhaps an example of organisations playing catch-up when it comes to artificial intelligence. Once again, the speed of the AI opportunity means organisational readiness is a long way behind the capability and availability of the technology.
‘Deploying AI: Balancing Power, Performance and Place’ – a Tech Monitor / AMD executive roundtable discussion – took place on Tuesday, 16 September 2025 at the Nimb Hotel, Copenhagen, Denmark.
