Theses: bachelor's and master's theses

07.01.2025, Abschlussarbeiten, Bachelor- und Masterarbeiten

Background and Context: Transparency in Artificial Intelligence (AI) is a cornerstone of responsible AI practices, ensuring stakeholders understand how AI systems are designed, trained, and deployed. As AI increasingly influences critical domains such as healthcare, finance, and law enforcement, transparency becomes essential for accountability, trust, and harm prevention. However, achieving transparency presents unique challenges due to the technical complexity of AI systems, proprietary restrictions, and conflicts with other responsible AI principles like privacy and intellectual property rights.

Project Goals: -Define Transparency in AI: Develop a clear and accessible definition of transparency, emphasizing its role in ensuring accountability, trust, and harm prevention. -Analyze Challenges: Examine why achieving transparency is difficult and how it may conflict with other responsible AI principles. -Explore Practical Applications: Investigate existing tools and methods for implementing transparency, such as model cards, dashboards, and explainable AI (XAI) toolkits. -Case Studies: Analyze examples of successful and failed attempts at achieving transparency in AI to provide practical insights. -Propose Solutions: Identify what is missing to make transparency effective in AI and recommend changes, including interdisciplinary collaboration and public education.

Requirements for student: -Strong interest in responsible AI practices and societal impacts of technology. -Background in AI, data science, or related fields (prior experience with transparency tools or concepts is a plus). -Strong writing (English) and analytical skills. -Ability to collaborate and meet project deadlines.

How to Apply: Interested Master’s students should send a brief CV and a 300-word motivation letter (optional but good to have) to dalia.ali@tum.de by 20 January 2025.

Kontakt: dalia@ali.tum.de