With the democratization of artificial intelligence and more recently the rise of Large Language Models (LLMs) such as ChatGPT, the interest in using AI or machine learning for security has increased.
Some possible thesis topics
- Using LLMs to establish contexts for cybersecurity events.
- Writing secure code using LLMs or analysing code for security vulnerabilities using LLMs.
- Testing web applications with regard to the OWASP top 10 Machine Learning Vulnerabilities [1].
- Automated offensive cybersecurity (social engineering, payload generation, and external knowledge gathering).
Other topics can be possible depending on interest.
[1] https://owasp.org/www-project-machine-learning-security-top-10/