Daryna Oliynyk

I am a PhD student at Vienna University of Technology, advised by Andreas Rauber, working on machine learning security, with a focus on the interplay between security and model compression.

I’m also conducting research on model stealing, which was the main topic of my master’s thesis. Alongside my doctoral research, I work as a security ML researcher at CDL AsTRa at University of Vienna and SBA Research.

Highlights

4th place at the European Cybersecurity & AI Hackathon Championship finals Jul 2026
Visiting the Dormant Neurons group at CISPA Jul 2026
From Dagstuhl to a preprint on agentic AI in cybersecurity Jun 2026
2nd place at the Vienna round of the European Cybersecurity & AI Hackathon Dec 2025
Invited talk at ESSAI 2025 Nov 2025
Winner of OCG Förderpreis 2025 May 2025

Selected Publications

2026

Rising From the Ashes: How Agentic AI is Unblocking Challenges in Cybersecurity

Gabriela F. Ciocarlie, Kathrin Grosse, Somesh Jha, and 3 more authors

2026

Under review

DOI Bib HTML

@misc{oliynyk2026rising,
  title = {Rising From the Ashes: How Agentic AI is Unblocking Challenges in Cybersecurity},
  author = {Ciocarlie, Gabriela F. and Grosse, Kathrin and Jha, Somesh and Oliynyk, Daryna and Paverd, Andrew and Wressnegger, Christian},
  year = {2026},
  note = {Under review},
  doi = {10.48550/arXiv.2606.23138},
}

2025

I Stolenly Swear That I Am Up to (No) Good: Design and Evaluation of Model Stealing Attacks

Daryna Oliynyk, Rudolf Mayer, Kathrin Grosse, and 1 more author

2025

Under review

DOI Bib HTML

@article{oliynyk2025stolenly,
  title = {I Stolenly Swear That I Am Up to (No) Good: Design and Evaluation of Model Stealing Attacks},
  author = {Oliynyk, Daryna and Mayer, Rudolf and Grosse, Kathrin and Rauber, Andreas},
  year = {2025},
  note = {Under review},
  doi = {10.48550/arXiv.2508.21654},
}

Attackers Can Do Better: Over- and Understated Factors of Model Stealing Attacks

Daryna Oliynyk, Rudolf Mayer, and Andreas Rauber

In IEEE Conference on Secure and Trustworthy Machine Learning (SaTML), Apr 2025

DOI Bib HTML

@inproceedings{oliynyk2025attackers,
  title = {Attackers Can Do Better: Over- and Understated Factors of Model Stealing Attacks},
  author = {Oliynyk, Daryna and Mayer, Rudolf and Rauber, Andreas},
  booktitle = {IEEE Conference on Secure and Trustworthy Machine Learning (SaTML)},
  year = {2025},
  month = apr,
  doi = {10.1109/SaTML64287.2025.00016},
}

2023

I Know What You Trained Last Summer: A Survey on Stealing Machine Learning Models and Defences

Daryna Oliynyk, Rudolf Mayer, and Andreas Rauber

ACM Computing Surveys, Jul 2023

DOI Bib HTML

@article{oliynyk2023survey,
  title = {I Know What You Trained Last Summer: A Survey on Stealing Machine Learning Models and Defences},
  author = {Oliynyk, Daryna and Mayer, Rudolf and Rauber, Andreas},
  journal = {ACM Computing Surveys},
  doi = {10.1145/3595292},
  url = {https://dl.acm.org/doi/10.1145/3595292},
  year = {2023},
  month = jul,
}