INSAIT’s Maria Drencheva discusses AI safety and security at Digitalk 2026

Maria Drencheva from INSAIT spoke yesterday at Digitalk 2026 about one of the most important challenges facing AI today—safety and security. Digitalk is an annual conference focused on technology, innovation, and digital transformation in Southeast Europe.

Maria discussed the opportunities and risks of agent-based AI, the threat large language models pose to online anonymity, their implications for mass surveillance, and the ethical challenges associated with the use of AI in national security, among other important topics.

At just 21 years old, and as a participant in INSAIT’s special four-year EXPLORER bachelor’s program aimed at developing talented Bulgarian students, Maria has already co-authored three research papers accepted at leading global AI conferences, including ICLR and ICML. She is also the first bachelor student in Bulgaria’s history to have a paper accepted at ICLR.

As a member of Bulgaria’s national mathematics team, Maria is a laureate of numerous international mathematics Olympiads, including a gold medal with a perfect score at the European Girls’ Mathematical Olympiad (EGMO) 2023. In 2025, she was a Quant Research Intern at Man Group, and in the summer of 2026, she has been selected for ETH Zurich’s Summer Research Fellowship programme.