Work­place AI in the fin­an­cial sec­tor

Analysis reveals hurdles to its use in everyday working life

Many companies expect enormous productivity gains from the use of artificial intelligence (AI) in everyday working life, so-called workplace AI (WKI), especially in knowledge-intensive areas such as consulting, administration, risk analysis, market research and customer acquisition. Although applications such as ChatGPT or Microsoft 365 Copilot have arrived in everyday office life, their potential often remains unutilised in practice. Business Informatics experts Prof. Dr Franz Strich and Prof. Dr Simon Thanh-Nam Trang from Paderborn University have systematically investigated the successful implementation of AP AI in eleven savings banks. The results of the survey show not only the hoped-for productivity gains but also various development potentials in the dovetailing of technology, employees, processes and strategy. In response to this gap, the Paderborn researchers have developed a new implementation model with twelve specific fields of action that supports companies in sustainably anchoring AP AI in their day-to-day work.

From trial and error to daily use

The study focussed in particular on the so-called "S-KIPilot", an AP AI specially developed for the savings banks. This revealed an interesting phenomenon: added value and innovation through AP AI first begins in the day-to-day work of employees. They initially test, adapt and integrate the technology independently into their daily tasks - often before an overarching strategic categorisation is carried out by the management level. Contrary to previous assumptions about the implementation of new technologies, the strategic guidelines of AP AI are less important for many employees. "Workplace AI is changing the way work is organised. For employees, the focus is shifting from pure system user to active architect. However, they also need the appropriate skills for this, for example to critically review AI results," says Professor Strich. Companies should therefore fundamentally rethink their control logic, their processes and their role models. This includes providing employees with targeted training, defining clear deployment scenarios and offering employees scope for the systematic integration of AP AI.

Otherwise, there is a risk of a productivity paradox: "Productivity gains through the use of workplace AI are cancelled out by increased control requirements or a lack of cultural embedding. Factors such as concerns about one's own job or the noticeably growing digital divide within the workforce are particularly visible," continues Professor Strich.

From utilisation to sustainable value creation

"The biggest mistake is to plan workplace AI first. AI works the other way round: introduce it, use it and only then manage it strategically," says Professor Trang. The introduction of AP AI therefore follows a new process: Firstly, employees test the technology in their day-to-day work ("run") and develop their own use cases ("build"). Strategic management ("Plan") only takes place in the third step. This approach to the introduction of new digital technologies, which differs from previous implementation strategies, presents many companies with relevant challenges.

While companies consciously create space for practical experimentation, a central task becomes visible in the strategic phase: organisations must make transparent where AI actually contributes to value creation - and where it does not. "Only by systematically recording and managing these effects will companies be able to move from experimenting with AI to strategic value creation in the long term," says Professor Trang.

The white paper is available free of charge to all interested parties [German only].

This text was translated automatically.