IT trends in research: universities focus on AI and cybersecurity

Contents

IT security and artificial intelligence are currently in vogue in the IT departments of universities and research institutions. This is the result of a survey conducted by the Centers for Communication and Information Processing in Education and Research (ZKI). This development is also reflected in the investments made by organizations, which are predominantly focusing on commercial AI applications. IT security also plays a role in current legislation. Other relevant topics include cloud computing, skills shortages and digitalization. German universities continue to pay attention to digital sovereignty.

Investments by universities support IT trends

The most important area of investment for institutions in research and teaching is cybersecurity. Around 23% of respondents stated that their organization had increased spending in this area. This includes threat management, security operations centers and compliance. Likewise, 13% have made additional investments in artificial intelligence and data processing. Just under 12 percent are allocating funds to cloud and infrastructure. Small universities with fewer than 5,000 students focus on IT security and infrastructure, while institutions with up to 30,000 students also invest in AI and cloud services. Larger universities are more likely to invest in digital sovereignty projects.

IT security is also important for universities and research institutes from a regulatory perspective. Around 14% of respondents stated this, pointing to the NIS2 directive, for example, which also requires training for managers. Around 11% emphasize the legislation surrounding data processing and protection. They mainly mention the GDPR, new data protection audits and Schrems III. In third place is the European AI Act, to which a good nine percent of respondents refer. Data protection is a high priority at small universities. As the number of students increases, universities are increasingly focusing on laws relating to the cloud and AI. Large universities are focusing on cybersecurity and institutional governance.

Commercial AI tools dominate at universities

The use of commercial AI applications such as ChatGPT predominates at universities and research institutes. They are used in almost one in four institutions. Institutional tools, such as HAWKI or bwGPT, are only used by around ten percent of respondents. Open source alternatives such as Mistral were mentioned even less frequently. The proportion here was just under seven percent. Small universities in particular, with fewer than 5,000 students, are increasingly relying on commercial applications and do not have their own AI infrastructures. As the number of students increases, so does the proportion of use of institutional applications and in-house infrastructures. Only at large universities with more than 30,000 students does the use of open-source language models and AI tools predominate.

Almost one in five research and teaching institutions is currently involved in setting up AI infrastructures and basic services. These include local language models, cloud services and GPU clusters. For 11 percent of respondents, the focus is on chatbots, which are to be used in areas such as student advisory services or IT support. Just under 10 percent are currently working increasingly on the use of AI tools in collaboration applications, such as Microsoft Copilot, for whose voice and reasoning chatbots there are no longer any usage restrictions. The tools are intended to support employees in the processing and translation of texts and in examinations.

Project management determines the success of digitalization

Various organizational units are responsible for digitalization at universities. In addition to a central IT center, they include strategic and operational units as well as units for the digitalization of administration, research and teaching. At small universities with fewer than 5,000 students, digitalization projects are usually initiated by the individual faculties. IT centers are mainly responsible for basic digital services and provide support for university members. Large universities with more than 15,000 students, on the other hand, tend to rely on central management by staff units. IT centers are also increasingly involved in digitalization projects here. They are also involved in the provision of new services.

Almost one in five respondents identified project management and coordination as a success factor in digitalization. More than 11 percent point out that difficulties in project management have led to failed digitization attempts. There needs to be as clear a structure as possible, as well as clearly formulated goals and schedules. Unclear ideas, a lack of competence and a lack of responsibilities are particularly obstructive. Sufficient resources are also important for the success of digitalization projects. Almost 14 percent see this as a success factor, while 13 percent see it as a reason for failed projects. In addition to financial resources, there must be sufficient staff with the necessary expertise. Frequent changes due to fixed-term contracts are also detrimental.

The ZKI association has a total of around 250 members. Most of these are research and teaching institutions that operate their data centers, but there are also managers who are responsible for operations. For the study, the authors surveyed around 350 managers from IT departments and data centers at universities and research institutes in the DACH region and France. In addition to the annual survey on current IT trends, the researchers examined the focus topics of AI and digitalization. Most recently, the Bundeswehr University in Munich was affected by a cyberattack.

(sfe)