Funding of Teaching project “BreAkIng bad – hacking the AI Act"

12 January 2023, by Chris Biemann

UHH/Regneri

The teaching project „BreAkIng bad – hacking the AI Act“ (designed by Dr. Michaela Regneri, together with Fabian Burmeister) was awarded  a 1-year funding within the teaching lab program by DDLitLab at UHH.

 
Course Summary
In the course, students develop consulting concepts for companies on the use of artificial intelligence (AI) and on compliance with the so-called "AI Act". In partnership with a consulting company, the project will be guided professionally and reality-based.

The AI Act is an upcoming EU regulation for the governance of AI. Currently, it is controversial how far the AI Act can fulfill its purpose to protect society. The consulting concepts developed by students will help to uncover weaknesses of the AI Act.

In an innovative hermeneutic teaching approach, students act as "Advocatus Diaboli": Consulting strategies are designed to show how the AI Act can be met with minimal effort while maximizing economic profit and accepting societal risk. The concept is similar to "ethical hacking" of software, where vulnerabilities are detected through elaborate but consequence-free attacks. Here, the partner company will show leeway in which they