Research Focus: Development of an intuitive workflow for AI-based image generation in the form of verifiable credentials

A verifiable credential is a digital certificate that is signed by the issuer and the holder using cryptographic methods to confirm its authenticity. These credentials contain information about a subject, who is usually the holder of the VC. This enables digital versions of various identity documents, such as student ID cards, to be created [2].

The characteristics of self-sovereign identity in the form of verifiable credentials can be clearly illustrated using the SSI interaction diagram shown in Figure 1. For example, if a university administration, acting as an issuer, issues a student ID card to a student in the form of a verifiable credential, this credential is stored exclusively on the holder’s (the student’s) device. If the student wishes to identify themselves to a library, for instance, as an accepting entity, they can present this credential. The library does not need to contact the university administration. The authenticity of the verifiable credential is verified by means of the digital signature with which the university administration has signed it. All that is required is the library’s trust in the university administration. The library can also check a trusted data register to see whether the credential has been revoked – for example, due to compromise – and reject it accordingly.

The Hidy app is an implementation of an EUDI wallet. It was developed as part of the ID-Ideal research project at Mittweida University of Applied Sciences, at the Blockchain Competence Centre Mittweida (BCCM). The project was funded by the Federal Ministry for Economic Affairs and Climate Action (BMWK) as part of the ‘Showcase for Secure Digital Identities’ innovation competition.

Apart from its use for digital identity documents, certificates and so on, technical considerations open up further possible applications. Within the ‘subject’ – which contains the data confirmed by the signature – there are few restrictions on data types or size. Binary data can also be incorporated in this way, enabling the embedding of any file format. This includes images. These possibilities have given rise to a desire to demonstrate that an EUDI wallet is also suitable for scenarios beyond identification and that entire workflows can be implemented using it. This has resulted in a software landscape that generates AI-generated images or empowers users to create AI-generated images and store them as Verifiable Credentials in the Hidy Wallet. The architecture of this is shown in Figure 2.

The ‘Stable Diffusion Proxy’ component makes the functions of the Stable Diffusion model available to the other components. This AI model is responsible for generating the AI images. ‘Art Bubble’ is the actual programme with which users interact by creating a prompt through selecting words. This can be seen in Figure 3. The QR code that appears in the top right-hand corner of the screen can be scanned using the Hidy Wallet. Users are then offered the image as a verifiable credential. If they accept this, the image is passed on to the ‘Shop Integration’ component. This creates a shop product for large-format prints via an external service provider. In the process, each image is given a title and a description using a large language model. An example of such a shop product can be seen in Figure 4.

Florian G. Hechler says of the research topic: “Working on this project is both very exciting and enlightening for me. I see this self-sovereignty in the field of digital identification as a major step forward in data protection. The coming years will show just how well the EUDI wallets are taken up by the public. I hope that their widespread adoption will be boosted by their extensive range of applications and that, together with my colleagues, I have made a positive contribution to the future.”

This article is based on: Hechler, F. G., Meisel, M. & Ittner, A. (2025). “Development of an intuitive workflow for AI-based image generation in the form of verifiable credentials”. In: Proceedings of the 25th Academic Talent Conference 2025, pp. 111–115. Available at: www.hs-merseburg.de/fileadmin/Forschung/Nachwuchswissenschaftlerkonferenz/NWK25/NWK25_Tagungsband.pdf (Accessed 15 July 2025).

Florian Gerhard Hechler is studying Applied Computer Science, specialising in IT security, at Mittweida University of Applied Sciences. Since March 2023, he has been working in the field of software engineering at the ‘Blockchain Competence Centre Mittweida (BCCM)’ at Mittweida University of Applied Sciences.

[1] Granc, F.; Fiedler, A. (2024): Towards European sovereignty through interoperability? Datenschutz Datensich 48, 217–221. doi.org/10.1007/s11623-024-1912-3

[2] Ehrlich, T.; Richter, D.; Meisel, M.; Anke, J. (2021): Self-Sovereign Identity as the basis for universally applicable digital identities. HMD 58, 247–270. doi.org/10.1365/s40702-021-00711-5