Cyberagentur contributes security perspective to white paper on generative AI for industry and companies
Dr. Daniel Gille from the Agentur für Innovation in der Cybersicherheit GmbH (Cyberagentur) has contributed as a guest author to the white paper “Using Generative AI responsibly” published by the Plattform Lernende Systeme. It shows how companies can use generative AI securely, legally compliant and strategically for digital sovereignty.
The Cyberagentur is contributing its expertise on the safe and confident use of generative artificial intelligence to the white paper “Using generative AI responsibly – impetus for companies and industry” published by Plattform Lernende Systeme (PLS). Dr. Daniel Gille, Acting Head of the Cyberagentur’s Key Technology Department and Head of Artificial Intelligence, contributed as a guest author.
The white paper classifies generative AI as a key technology for companies, industry and administration. It analyzes opportunities for efficiency, knowledge management, software development and production. At the same time, it identifies the risks that can arise from uncoordinated and unsecured use. These include hallucinations, data leakage, new attack surfaces, dependencies on large platform providers and unresolved questions regarding governance, compliance and responsibility.
From the Cyberagentur’s perspective, this debate is relevant to security policy. Generative AI is not just a productivity tool. It is changing digital value creation, software development, security architectures and decision-making processes. It therefore directly affects the technological sovereignty of Germany and Europe. Companies must therefore clarify early on how they want to use generative AI as users and, if necessary, develop their own specialized and more controllable solutions.
A particular focus of the white paper is on Small Language Models, or SLMs for short. These smaller language models require less computing power and can be adapted for domain-specific tasks and often operated locally. This results in advantages for data protection, data security, cost control and digital sovereignty. SLMs can be a strategic alternative to global cloud LLM offerings, especially for sensitive industrial applications, public authorities and security-critical processes.
Dr. Daniel Gille: “Generative AI only offers a strategic efficiency advantage if security, traceability and sovereignty are considered from the outset. Companies don’t just need fast applications. They also need to be able to make reliable architectural decisions. Otherwise, new dependencies, new attack surfaces and a loss of control over data and value creation will arise.”
The white paper makes it clear that the use of generative AI requires clear guidelines. These include security by design, verified training and operating data, transparent responsibilities, technical access restrictions, robust AI governance and continuous quality control. Equally important is the training of employees. This is the only way to exploit opportunities without weakening security, data protection and compliance.
The white paper provides practical guidance for companies. SWOT analyses show examples of how generative AI can be evaluated in knowledge management, industrial applications and software development. The aim is to systematically identify potential, recognize risks at an early stage and derive viable strategies.
Dr. Daniel Gille’s involvement underlines the Cyberagentur’s commitment to bringing together research, application and security needs at an early stage. Generative AI must be designed in such a way that it does not create new vulnerabilities, but instead enables robust, verifiable and sovereign digital structures.
Further information and registration: