The results of the KPMG study "Generative AI in the German economy":
- 53 percent of companies want to increase investments in generative AI in the next 12 months, half of them by 40 percent or more (55 percent)
- 67% each expect an increase in turnover and automation, 65% expect costs to fall, 55% hope for increased productivity
- 37% fear negative effects; the top three effects include job losses, data protection problems and a lack of employee acceptance
- Need to catch up in terms of strategy, governance and training: Only just under one in three companies have a strategy in place; 8 percent have a complete governance model and 63 percent do not consider themselves well prepared for the EU’s AI Act; only 38 percent of respondents consider their company to be well or very well prepared to train employees in dealing with Generative AI
Generative AI is becoming increasingly relevant for German companies: More than half of companies (53%) are planning to significantly expand their investment activities. More than half of these companies even plan to increase their investments by 40 percent or more. Although the topic of generative AI is being approached more strategically and initial use cases are being implemented, there is still room for improvement: 31% of the companies surveyed have drawn up a strategy, while 49% are planning such a strategy or have already started to implement it. More than one in three companies already has specific use cases in place, while 35% intend to implement use cases in the next 12 months.
These are the findings of the study “Generative AI in the German economy”, for which KPMG in Germany surveyed over 280 decision-makers from the German economy in March 2024. “The phase of experimenting with generative AI is slowly giving way to a phase of implementation. Companies are taking a more strategic approach to generative AI and using the technology on a day-to-day basis. But there is still room for improvement,” says Benedikt Höck, Partner at KPMG AG Wirtschaftsprüfungsgesellschaft. The top three most common AI models currently in use include Google Gemini (34%), GPT 3.5 / 4.0 (34%) and the German player on the market Aleph Alpha (13%). 36 percent of companies do not yet have a model in use.
34 percent
Google Gemini
34 percent
GPT 3.5 / 4.0
13 percent
Aleph Alpha
Overview of the top three most common AI models currently in use (source: KPMG)
Training and governance are challenges
Overall, companies expect a number of positive effects from the use of generative AI. Above all, respondents expect an increase in turnover (67%), a higher degree of automation (67%) and lower costs due to greater efficiency (65%). On the other hand, the survey participants also see negative effects. In each case, 37% of them cite job losses, a lack of employee acceptance and problems with data protection.
Companies also see challenges in employee training and in the areas of governance and regulation. Although 55% of respondents believe “strongly or very strongly” that employees will have to adapt their way of working to use generative AI effectively, only 38% of them believe they are “well or very well” prepared to train their employees in the use of generative AI. Only 8 percent of companies have a complete governance model including a committee for managing AI. 63 percent of companies also do not consider themselves well prepared for the effects of the European Union’s Artificial Intelligence Act.
Source: KPMG press release, May 13, 2024
