In companies in Mainfranken, AI has become part of everyday working life, but not in automated processes. This is shown by an online survey conducted by the Würzburg-Schweinfurt Chamber of Industry and Commerce and the University of Würzburg.
“When people in Mainfranken talk about artificial intelligence, it’s rarely about robots and almost always about the desk.” This is how Professor Carolin Wienrich, Chair of Computer Science VI (Psychology of Interactive Systems) at Julius-Maximilians-Universität Würzburg (JMU), summarises the picture that emerged from a survey of 53 companies conducted in collaboration with the Würzburg-Schweinfurt Chamber of Commerce and Industry.
AI has long been widespread in these companies; on average, 3.7 different tools are in use. ChatGPT is the most frequently mentioned, at around 62 per cent, followed by Claude and Microsoft Copilot at 45 per cent each, and the translation service DeepL at 36 per cent. Around a third of respondents use an in-house AI solution.
The applications are less spectacular than the debate might suggest. Research and writing texts are tied for first place at 91 per cent each, followed by summarising information at 66 per cent.
As soon as the focus shifts to deeper aspects of business processes, the figures dwindle: only 19 per cent cite process automation as an area of application. “AI has arrived in the workplace, but not yet in business processes,” concludes Julia Holleber, digitalisation officer at the Chamber of Industry and Commerce (IHK).
This is consistent with the role respondents attribute to the technology. The most frequently chosen image was that of a personal assistant who helps with organisation and routine tasks (32 per cent). This is followed by the administrative assistant, who independently carries out clearly defined tasks (17 per cent), and the project partner (15 per cent). Only 13 per cent view AI as an equal or superior. It is striking that those who believe AI can act as an independent partner or colleague also rate their own knowledge higher than those who see AI primarily as an assistant.
Scepticism is not a knowledge gap
Overall, confidence prevails amongst companies when it comes to AI. On a scale of 1 (‘major risk’) to 10 (‘major opportunity’), the average score is 6.8, with just over half (43 per cent) awarding 8 points or more. 30 per cent of respondents tend to see AI as posing risks to their own professional development. This group differs little from the rest in terms of their AI knowledge, so their reluctance cannot be explained by a lack of knowledge.
This is also reflected in the question of which skills will matter in the future. Just over half of those surveyed cite the ability to critically assess and interpret AI results. Data protection, legal issues and ethical responsibility follow, whilst purely technical skills lag behind.
Almost half of companies have rules
When it comes to internal regulations, companies are further ahead than one might expect: almost half (49 per cent) have a written policy or works agreement, whilst 30 per cent expressly prohibit the use of private AI applications. At the same time, around a quarter of those surveyed (23 per cent) have so far been operating without any regulations whatsoever.
The difference is linked to company size: amongst companies with more than 50 employees, just over 70 per cent have a policy, whilst amongst those with fewer than ten employees, none do. The data shows, however, that clear rules do not hinder usage, as companies with a policy use AI more widely and intensively on average than those without one.
Consequently, there is a demand for practical guidance. For instance, 70 per cent would like to see real-world application examples, 53 per cent want changes to work processes, and 47 per cent require a legal framework. Basic knowledge of how AI works comes at the bottom of the list, at 38 per cent.
The need arises within the organisation
The survey provides a surprisingly clear answer to the question of exactly when AI training is needed: namely, when new digital or AI-supported processes are introduced within the company. By contrast, the traditional stages of a career – starting out, returning to work after a break or taking on a new role – all rank in the middle of the list.
Both newcomers and those with more experience recognise this high demand. More than half of those surveyed feel they are further ahead in their personal development than their own company. The problem lies less with the people than with implementation within the organisation. When it comes to course formats, preferences are very specific: around four-fifths want courses lasting no more than half a day, preferably online with a trainer or as guided self-study.
Survey reveals trends
The online survey took place in the summer of 2026. Fifty-three companies from Mainfranken took part, predominantly from the sectors of industry, IT and retail. Given the number of participants and the voluntary nature of the survey, this is a qualitative study that highlights trends but does not provide representative figures for the regional economy.
