The use of artificial intelligence has become part of daily business in European and global companies, yet widespread deployment frequently fails due to outdated IT structures. According to the global Cloudera study titled "The Great AI Re-Architecture" published on August 11, 2026, 77 percent of surveyed companies actively use AI applications. However, 95 percent of enterprises had to postpone or completely cancel planned AI initiatives over the past year. The primary causes cited include legacy data architectures, inadequate data governance and complex regulatory compliance requirements.
To resolve these blockades, many organizations are now planning a fundamental modernization of their IT systems. In the Cloudera survey, 72 percent of executives stated that they need to thoroughly overhaul their existing data infrastructure. The industry trend is moving clearly toward hybrid data architectures. Such models allow enterprises to keep business-critical data on-premises while flexibly connecting it to modern cloud-based AI models.
This discrepancy between strategic goals and technical reality is also reflected in the manufacturing sector across the DACH region. A survey presented on August 3, 2026, at the IFS Connect DACH conference among 91 executives from Germany, Austria and Switzerland illustrates the current state of implementation. The data shows that 43 percent of industrial companies are currently in the pilot phase. Only 27 percent use AI systems productively in individual business units, while merely 4 percent have achieved a company-wide rollout.
Despite the hesitant adoption rates, the practical benefits of automated systems are already tangible across corporate operations. Four out of five companies, representing exactly 80 percent of respondents, report concrete time savings and significant efficiency gains from AI deployment. At the same time, 55 percent of respondents identify a lack of data access and poor data quality as the main bottleneck. These deficiencies frequently prevent the successful transition from initial test projects into regular production.
In addition to internal data issues, regulatory requirements such as the transparency obligations under the EU AI Act are increasing pressure on businesses. Operational liability remains with operating companies, which is why financial approval limits and human-in-the-loop processes are becoming standard for autonomous AI agents. Furthermore, an investigation by netzpolitik.org on August 10, 2026, revealed that government plans to double German data center capacity by 2030 lack specific data on future water consumption. Overcoming technical, legal and infrastructural hurdles is therefore becoming the decisive factor for future AI initiatives.