Economic researchers and industry associations are painting a nuanced picture of artificial intelligence adoption across business sectors. While corporate leaders widely recognize the strategic importance of automated systems, organizations are still struggling to fully exploit these capabilities. Recent findings from the Munich-based ifo Institute, German digital association Bitkom, and the National Bureau of Economic Research illustrate this dynamic. Optimism remains strong across industries, but transforming technological access into tangible output requires substantial process restructuring.
According to a survey conducted by the ifo Institute led by Dr. Klaus Wohlrabe, roughly 70 percent of surveyed companies in Germany expect noticeable productivity increases from deploying AI. Over a five-year horizon, businesses anticipate average productivity gains ranging between 8 percent and 16 percent. However, Wohlrabe pointed out that this broad corridor reflects deep underlying uncertainty. While many executives grasp the general promise, they still struggle to measure the concrete operational impact on their daily workflows.
Sectoral differences remain pronounced across the German economy. The ifo data shows that the service industry holds the most optimistic outlook, expecting productivity increases between 10.2 percent and 20.9 percent, followed closely by wholesale and retail trade at 9.1 percent to 23.0 percent. In manufacturing, projections range between 7.8 percent and 17.3 percent. The construction sector expects the lowest impact, with gains estimated between 8.2 percent and 14.8 percent, largely due to the physical and custom nature of job site work.
A survey by Bitkom among 602 German companies provides a grounded assessment of current adoption maturity. Around 73 percent of enterprises categorize AI as the most crucial technology for their future competitiveness, and only 26 percent dismiss it as an overhyped trend. However, practical execution lags behind strategic intent. Even though 57 percent of businesses now use AI tools, 59 percent of those active users admit they are not tapping into the full potential, and only 3 percent report extensive exploitation of capabilities.
Concrete macroeconomic evidence on day-to-day work effects comes from a National Bureau of Economic Research working paper authored by A. Bick, A. Blandin, and D. Deming. Their analysis indicates that between 1 percent and 5 percent of all corporate working hours are directly assisted by generative AI. On an aggregate level, this translates into a labor time savings equal to 1.4 percent of total working hours. Furthermore, 23 percent of surveyed workers use generative tools at least weekly, with 9 percent relying on them daily.
In a historical perspective, the NBER researchers note that generative AI adoption is spreading as fast as the personal computer did and outpaces the early adoption curve of the internet. The current discrepancy between high hopes and modest measured time savings marks a classic transition period. For enterprises, the focus must shift from disconnected pilot projects toward structured workforce upskilling, clear governance guidelines, and deep process integration.