• Top management in Germany expects clearly positive EBIT effects from AI – while operational levels and international peers are significantly more skeptical
• Security concerns, lack of market transparency, and skills gaps are slowing down value creation from AI

German companies are often optimistic about the impact of their AI investments on financial performance. According to the AI study “From AI to EBIT” by Horváth, 75 percent of executives at large companies in Germany expect a strongly positive impact of their AI investments on earnings before interest and taxes (EBIT). Globally, this figure stands at only 54 percent. Notably, senior leadership – board members and executive teams – is significantly more optimistic than operational management. While three-quarters of surveyed CxOs worldwide expect a strong positive EBIT impact from AI investments, only 21 percent of department-level managers share this view.
“This gap in expectations clearly indicates that strategy and implementation are often misaligned,” says Matthias Emler, Partner at Horváth. “Expectations at the top management level continue to rise – making disappointment almost inevitable.”

AI hype fuels demand—and poor decisions
In four out of five German companies, top management’s willingness to invest in AI-based solutions is currently higher than for other IT investments. As a result, vendors are aggressively capitalizing on this momentum. A majority of respondents report this trend (Germany: 79 percent; U.S.: 84 percent). Consequently, executives regularly approach IT departments to discuss acquiring new AI solutions – often without thoroughly assessing their actual necessity or business value. Survey results indicate a high risk that purchased solutions ultimately fail to integrate optimally with existing systems.
“Driven by high expectations from senior management and strong sales pressure from vendors, AI solutions are sometimes purchased and implemented without a systematic selection process,” Emler explains.
Companies are also increasingly observing that the price-performance ratio of many AI solutions is becoming unbalanced due to the hype. Seventy-nine percent of respondents report this perception. The same proportion notes that available AI applications often fall significantly short of expectations in terms of maturity and functionality.

Lack of transparency and understanding
A key driver of “mis-purchases” is a lack of transparency, insufficient market overview, and limited understanding of technology at the management level. Eight out of ten respondents in Germany state that a significant share of managers struggle to fully grasp new AI technologies. Additionally, seven out of ten companies cannot clearly identify which AI applications are in use across the organization.
“The boom in AI tools – combined with continuous investments by individual departments – makes it increasingly difficult to maintain an overview,” says Horváth Partner Matthias Emler.

AI security risks particularly concern German companies
Seventy-four percent of respondents view the use of non-approved AI applications as critical. As the proliferation of AI tools increases within organizations, cybercriminals are also leveraging the technology more intensively. Eighty-four percent of decision-makers report that AI-driven cyberattacks are becoming more sophisticated. Sensitivity to AI-related cybersecurity risks is more pronounced in Germany than in the U.S. (10–15 percentage points higher).

Concerns about declining professional expertise
In 81 percent of surveyed German companies, management is concerned that employees increasingly rely on AI tools such as ChatGPT and may process outputs without sufficient verification or blindly trust generated information.
“This risk is real,” says Emler. “It is therefore essential to complement the systematic use of AI with a company-wide training and enablement program that conveys how the technology works, its risks, and corporate guidelines.”
If AI transformation is supported by such programs, professional expertise and AI capabilities can be effectively combined – leading to more efficient and higher-quality outcomes.

About the Study
For the study “From AI to EBIT – How AI Finally Delivers Business Value,” Horváth surveyed more than 200 companies and organizations across the DACH region, the United States, and the Nordics, each with annual revenues of at least €200 million. The German sample alone includes more than 100 companies.