AI is reshaping software development as we know it. With just a few clicks, anyone with an idea can now generate advanced applications. However, the rise of “vibe coding” also introduces new security risks. Code suggestions from language models can result in vulnerabilities and data leaks, particularly in small and medium-sized enterprises with limited resources.
The project “Understanding and Supporting Secure AI Tools in Software Development Teams” aims to develop guidelines and tools that enable developers to use AI in software engineering safely and responsibly. By doing so, it directly strengthens cybersecurity and ensures that AI supports quality as well as productivity.
Support for this initiative reflects DIREC’s commitment to driving research that delivers concrete value to Danish companies while addressing society’s need for secure and responsible digitalization. With its focus on cybersecurity and the challenges faced by SMEs, the project will play a key role in strengthening Denmark’s technological resilience and advancing the political vision of a secure digital future.
The project is rooted in a strong partnership between the IT University of Copenhagen, the University of Southern Denmark, and three Danish software companies: Mjølner Informatics, CyberJuice, and Clockwork Consulting. Through embedded research, co-design, and workshops, the researchers will combine academic expertise with industry practice to create practical solutions that accelerate responsible technology development.
The project will result in the development of an AI fact-checker that helps developers assess the security of AI-generated code in real time. In addition, the team will produce practical guidelines for software teams integrating AI into their workflows. These outcomes will address both social and technical dimensions—for example, how AI reshapes team dynamics, and how developers can best prevent errors and security breaches.
 |
Louise Barkhuus – Professor – IT University Copenhagen
|
 |
Adam Alami – Associate Professor – University of Southern Denmark
|
 |
Elda Paja – Associate Professor – IT University Copenhagen
|
 |
Oksana Kulyk – Associate Professor – IT University Copenhagen
|
 |
Peter Mayer – Associate Professor – University of Southern Denmark
|
This project is particularly relevant for CyberJuice because it aligns directly with our core vision: to empower SMEs with cutting-edge, actionable security knowledge. As a fast-growing company, we face the same challenges many others do, that is navigating the evolving landscape of AI tools, security risks, and the human factors in software development. Participating in this project will not only help us enhance our own processes but also contribute to industry-wide understanding of AI’s role in secure software development.
Bodil Biering
CEO and Founder
Cyber Juice ApS
By joining this project, we would like to get more evidence of what works and in what context. The huge amount of automatic software tools currently available shows, that there is no consensus on how to ensure high quality software. We see this project as an opportunity to gain more knowledge and perhaps contribute to consolidate metrics and methods.
Jens Bæk Jørgensen
Senior Programme Manager
Mjølner Informatics
We believe the type of knowledge that this project will provide can help companies like ours better understand how to potentially use AI (or not) in security and safety related software development. As a small company, we do not have resources to invest in expensive tools and investigations that would uncover where in our process AI tools would be relevant, and as such, this research is very relevant to our software production. By contributing our practical expertise and participating in the project, we can help shape a tool that helps across the software development industry.
Tobias Jørgensen
Founding Partner
Clockwork Consulting A/S
