Collaboration between a global verification and anti-fraud leader and renowned German university advancing research and education to tackle AI-powered digital fraud
We are announcing a pioneering research partnership with Constructor University, a distinguished German institution renowned for its interdisciplinary approach. This collaboration marks the first key initiative within our newly launched AI Academic Program – a strategic effort to advance innovative, AI-driven solutions that address the growing threats to democratic institutions and digital identities worldwide, while strengthening the global fight against synthetic fraud.
Constructor University is one of Germany’s top private research institutions, known for its international focus. Professor Dmitry Vetrov, a leading AI researcher in Bayesian methods, generative models, and deep learning with alumni contributing to startups like Deepseek. Complementing this, Professor Hilke Brockmann leads the AIDE (AI, Democracy and Education) project, examining AI’s social and ethical impacts on trust and governance. Together, their insights strengthen the partnership’s mission to develop transparent and effective AI-driven fraud prevention.
With our co-founder and CTO, Vyacheslav Zholudev, who earned his doctorate at Constructor University, the partnership serves as a way for the company to give back to the university while advancing research and talent development to address AI-powered fraud globally.
We will make a direct investment — funding academic research, scholarships, and open innovation — to accelerate a shared mission in AI safety and global talent development. Further, students will be placed within our product and technical teams. These experiences will give emerging AI experts the chance to work on advanced anti-fraud tools, gaining practical skills and insight into real-world challenges in identity verification and digital security. Together, we aim to produce academic research and articles based on the collaboration, as well as open-source models shared publicly, to equip new generations to address emerging AI-powered threats in an increasingly digital world.
AI fraud today is fueling a complex crisis affecting many areas of society, including finance, healthcare, government services, ecommerce, social media, and more. According to internal data, global deepfake fraud surged 700% in Q1 2025 compared to the same period last year, while synthetic identity document fraud jumped nearly 200% across Europe and an alarming 577% and 378% in Germany and worldwide, respectively. These threats are eroding public trust, enabling financial crimes, and compromising privacy on an unprecedented scale.
“To fight deepfakes effectively, it is crucial to form alliances with top academic minds, which is why this collaboration is so important – no other player in the market is currently investing in this kind of research at such scale,” said Pavel Goldman-Kalaydin, Head of AI and ML at Sumsub. “Over the past two years, deepfake fraud has been scaling rapidly both in quality and quantity, and with the democratization of AI tools.This partnership will combine research and practical solutions to help society stay ahead of these threats. We’re honored to collaborate with the brilliant minds at Constructor University and hope this initiative encourages other leading universities to join our Academic Program, advancing research and education to combat AI fraud.”
“This collaboration brings together academia and industry to tackle some of today’s most pressing challenges. We are excited to work with Sumsub to explore these challenges through thorough academic research and practical work,” said Dmitry Vetrov, Professor of Computer Science at Constructor University. “Combining our diverse academic expertise with Sumsub’s leadership, we aim to create AI solutions that combat fraud and promote transparency, fairness, and trust in digital systems affecting millions.”
The deceptive nature of deepfakes – hyper-realistic but fake digital representations – makes them hard to detect, requiring constant vigilance and strong defenses. Our AI team is also advancing academic research, with peer-reviewed work by Viacheslav Pirogov accepted at ICML 2025 (International Conference on Machine Learning). His papers on real-world deepfake detection and zero-shot vision-language models highlight Sumsub’s role as both a technology provider and a contributor to global research. By combining academic and industry expertise, this partnership aims to close the growing gap between new digital fraud risks, current defenses, and public awareness.
Looking ahead, we plan to forge additional partnerships with leading academic institutions worldwide, reinforcing its ambitions to lead the fight against evolving digital fraud.
