Making AI Tangible and Accessible for SMEs

Artificial intelligence is a key driver of digital transformation, particularly for small and medium-sized enterprises (SMEs). But what potential does this technology truly hold, and how can challenges such as a lack of knowledge, data management issues, and legal uncertainties be addressed? 

In this eco Association interview, Dr. Christian Temath, Managing Director of the KI.NRW competence platform, and Andreas Weiss, Managing Director of the eco Association, discuss the specific opportunities that AI presents for SMEs. They emphasize that AI can only reach its full potential when supported by strong digital infrastructure, an innovation-friendly corporate culture, and practical support services.

Most SMEs have already mastered the ‘first wave’ of digitalization. Mr Temath, where do you see the greatest potential for AI to further develop these digital processes?

Christian Temath: In principle, there is exciting potential in almost all industries – for example in the creative sector, in retail and skilled crafts or in the service sector. In addition to the many AI tools already available today, new tools will be available to assist us in many areas of work in the future. Analytical and generative AI go hand in hand here: I see artificial intelligence primarily in product development and improvement: here, for example, AI will make suggestions for optimizing existing designs. 

AI can also be used in a variety of ways in manufacturing, for example in AI-supported quality assurance. Or in the energy sector for monitoring power grids. In addition, repetitive administrative processes in public administration bodies or in the medical sector, in particular, will be largely supported by artificial intelligence in the future – for example in the analysis or creation of documents such as doctor’s letters.

There is enormous potential here to overcome the shortage of skilled workers. But there are already a large number of tried-and-tested solutions available today, and there will certainly be more to come. AI can also provide support in the area of sustainability, whether in building energy management, CO₂-saving order and route planning or the on-demand manufacturing of fresh food.

The eco Association is at the forefront of the Gaia-X project. Mr Weiss, how do such projects contribute to providing trustworthy and safe AI solutions for SMEs that also meet data protection requirements and European standards?

Andreas Weiss: Basically, we are addressing the development of cooperative data ecosystems and the Gaia-X initiative has done groundwork when it comes to enabling trustworthy decentralized data collection and processing. For SMEs, this means access to AI solutions based on European data protection and security standards. Interoperable and resilient cloud infrastructures enable companies to use their data securely and connect it seamlessly with AI applications. Data sovereignty is crucial here, as it ensures that companies retain control over their data and can use it in a trusted environment. Concrete connectivity initiatives, such as Manufacturing-X, not only promote the deployment of AI and other value-adding processes, but also the trust of companies and their customers in these digital technologies.

Where do you see the greatest need for SMEs when it comes to artificial intelligence? Why is AI education and training so important?

Christian Temath: Developments are rapid and the number of AI offerings and products is increasing daily. However, as artificial intelligence is something completely new for many people and companies and SMEs in particular often lack the necessary knowledge, it is essential to make AI an important topic in education and training. There is a need for employees to learn and understand how AI works in principle. They should also include answers to the question of where AI can and cannot help in specific use cases. At this point, it is important for me to emphasize that AI should support employees, but not replace them. For important and essential topics, it is still the human being who makes the decision. Training and further education also discuss how artificial intelligence can be integrated into existing processes and the corporate culture. A frequently mentioned keyword in this context is the trustworthiness and safety of AI applications. There is still a certain degree of mystification and uncertainty here.

How can we accelerate the adaptation of AI in SMEs and create broad acceptance for it?

Andreas Weiss: My view is similar to Mr. Temath’s, but I would like to add that, in addition to traditional training and pilot projects, industry-specific training formats that specifically take into account the particularities of individual industries can play a decisive role. Particularly in sectors such as skilled crafts, logistics or the healthcare industry, there are specific challenges that require a tailored approach.

With its “AI in Practice” initiative, eco offers a central platform that provides companies with targeted support in gaining an overview of AI trends and developing manageable concepts for practical use. The initiative comprises various working groups that deal with data ecosystems, infrastructures and legal frameworks, among other things. Particular attention is paid to SMEs, for which practice-oriented tools, service catalogs and learning programs are provided.

In addition to this, we are strengthening the acceptance of AI in Germany with our “AI Future Tech” campaign, for example, where we give companies from a wide range of industries a forum to report first-hand on their best practices and AI flagship projects. In this way, we want to illustrate the practical use of AI as a catalyst for innovation and encourage companies to engage in exchange and initiate their own projects.

Another approach is to actively involve employees in the design and development of AI applications. This could be done via workshops, for example, in which specific use cases are developed within the company. Such formats not only create knowledge but also promote acceptance of and trust in AI solutions. We offer a series of training courses to become an ‘AI Officer’.

In addition, it is important to establish a culture of innovation that supports an open approach to new technologies. eco actively contributes to breaking down fears of contact and networking companies with each other in order to exploit synergies through events, specialist groups and working groups. Initiatives such as “AI in Practice” offer a space for the exchange of best practices and promote the development of concrete use cases – a decisive step towards bringing AI into the mainstream and strengthening the competitiveness of SMEs in the long term.

How exactly can KI.NRW provide support in such cases?

Christian Temath: This is where we come in with educational offers and support SMEs and start-ups in our federal state North Rhine-Westphalia (NRW), in particular, as they embark on their AI journey. These include, for example, our AI.Consultation, but also the AI.Kick-off presentations. We also offer the workshop format AI.Shadowing. Here, our experts go directly into companies and observe processes and workflows in order to identify and evaluate AI potential and develop relevant AI use cases together with the company. In the AI Design Sprint™, we develop concrete AI solutions together with the employees. Take a look at our digital AI.Map, which has more than 1,300 entries and maps the AI ecosystem in NRW. In addition to research projects, AI users and AI providers, it also includes courses of study and professional development opportunities that companies can use for their employees.

What kind of SMEs are approaching you and how do they become aware of you?

Christian Temath: We have noticed that there is a broad interest in AI. Companies of all different sizes and sectors are turning to us. These can be small craft businesses, an expert’s office, a medium-sized publishing house or industrial companies. They find out about KI.NRW and our providers through industry events organized by the various chambers of industry and commerce, for example, where we take part with introductory presentations or information stands. The initial impetus to get involved with AI often comes from tech-savvy employees, for example. And for some time now, more and more managers have also been keen to find out what AI has in store for them and what opportunities artificial intelligence offers their company due to the high level of media interest in AI.

Where do you see typical ‘mistakes’ that SMEs make when implementing AI? In your view, are there self-made ‘pitfalls’ that these companies could avoid in advance?

Christian Temath: Some companies have exaggerated expectations that artificial intelligence will solve all their problems. That is not the case. Rather, it is important to start simple and measurable with the first AI use case. You should start with a specific process, one that you understand very well. Because if you know exactly where the sticking points are, you can also make a good assessment of which adjustments you need to make with AI, for example, in order to improve a process. This could be optimized quality control in manufacturing or more efficient administrative processes.

A prerequisite for all corporate AI projects is suitable data with which to train the AI. In other words, the database must be right for AI to deliver good results that make sense in the context of a company. It is also important for the successful deployment of AI that AI is understood as a team sport in which employees are intensively involved in finding AI solutions. This is especially true for those who are to work with AI themselves in the respective processes or areas. This allows them to contribute their knowledge in order to jointly define a use case. This creates a positive atmosphere and acceptance.

What distinguishes NRW as an AI location? What is going well in NRW, and where is there room for improvement?

Christian Temath: We have many large companies, SMEs and a lively start-up scene in NRW – even with a unicorn: the AI translation tool DeepL – and many “hidden champions” in a wide range of industries. They are already actively using AI. We also have an excellent research landscape with universities, colleges and research institutions that work closely with start-ups and companies. For example, the OpenGPT-X project in NRW, led by Fraunhofer IAIS and IIS, has developed a large language model with a focus on European languages, which has been available as open source since fall 2024. A lot is also happening in the Rhine region, for example through Microsoft’s investment in AI data centers. And then there is the flagship initiative funded by the state government, for which we as KI.NRW are the patrons.

These are pioneering AI projects from various sectors such as retail, manufacturing, skilled crafts and healthcare. Experts from research and practice are working on solutions for real-world problems. This shows that NRW can do AI. But of course, there is always room for more of everything, for example, for significantly more SMEs to use artificial intelligence. There is still some catching up to do here. At this point, I would like to encourage companies to get involved with AI and simply get started. We are happy to provide support.

What else do you think is needed for AI to become a success story in Germany?

Andreas Weiss: We need strong infrastructures! Powerful data centers are the backbone for the development and application of modern AI models, which require enormous computing power and storage.

 The current eco study “Spillover effects of data centers: backbone of the AI revolution in Germany“ impressively shows that companies that rely on cloud and data center infrastructures are not only more innovative, but also achieve higher productivity gains. It is particularly striking that the proportion of revenue generated by new products and services is significantly higher for data center users than for companies without this infrastructure. The combination of strong infrastructures and AI opens up new opportunities for companies to develop digital products and services faster and more efficiently.

To achieve this, we also need a strong ecosystem of talents who are involved in AI development technologies, operations and adaptation in companies. Together with the German AI association, KI Bundesverband, we are currently planning the LEAM (Large European AI Models) project in the Rhineland region, which addresses precisely this. In other words, the creation of special AI models, the establishment of start-ups and innovative companies and the transfer from research to industry. This is also a significant value contribution for the structural change from coal mining to digital technologies.

These measures are supported by so-called IPCEIs (Important Projects of Common European Interest). A current measure known as “8ra” (pronounced oora or IPCEI-CIS) focuses heavily on cloud and edge infrastructures and enables flexible data collection and processing in the field, which also forms the basis for connected AI value creation processes.

What legal and regulatory challenges do you see in introducing AI in SMEs, and how can policymakers provide support?

Andreas Weiss: One of the biggest challenges for many companies is the uncertainty surrounding legal requirements, such as those defined by the EU’s AI Act. For SMEs in particular, it is essential that regulatory requirements are clear and can be implemented without creating unnecessary hurdles. Policymakers can provide support here by providing practical guidelines and funding programs that facilitate the implementation of AI. At the same time, international competitiveness and data sovereignty are important issues that should be promoted through coordinated European regulations.

As the Association of the Internet Industry, eco plays a central role in this by bringing the interests of our members into the political discourse and acting as an interface between business and politics. We support our members with well-founded information and are committed to ensuring that the legal framework is designed to be innovation-friendly and practical. Initiatives such as 8ra, Manufacturing-X and Gaia-X are examples of how cooperation at the European level can create and coordinate standards that meet regulatory requirements and promote data sovereignty and economic development.

The Framework Conditions working group of eco’s “AI in Practice” initiative have published a white paper on the AI Act. This provides concrete assistance for companies to better understand and implement the requirements of the Act.

Andreas Weiss is Managing Director of eco – Association of the Internet Industry. He has been with the association since 1998 and headed the E-Commerce and Logistics Competence Group and later the E-Business Group. Since 2010, he has been Director of EuroCloud Deutschland_eco e.V. and since 2019 a member of the board of Trusted Cloud, where he has been responsible for the association's cloud activities. Weiss is also involved in projects related to artificial intelligence, data protection and the GDPR, as well as the security and compliance of digital services in eco's Digital Business Models division.

Andreas Weiss is involved in a number of initiatives related to designing digital ecosystems, applying sovereign digital identities, and building AI operating structures with cloud and edge concepts, all with a focus on supporting the EU's ability to shape digital innovations and processes.

He started his freelance software development career in 1984 while still studying natural sciences. He coordinated international projects in service management for Hewlett-Packard and worldwide installation services for Sun Microsystems servers.

Dr Christian Temath has been the managing director of the competence platform KI.NRW since September 2020. KI.NRW is the central contact point for artificial intelligence (AI) in North Rhine-Westphalia. At the Fraunhofer IAIS in Sankt Augustin, he and his team are working on establishing the "AI made in NRW" brand and strengthening the technological sovereignty of North Rhine-Westphalia. With a PhD in Business Informatics, he has many years of experience in management consulting in the field of technology as well as in the practical application of AI technologies at an international e-commerce company.