An Italian cybersecurity startup used advanced artificial intelligence models to identify serious vulnerabilities in Apple systems and has now raised €2.1 million to turn that methodology into a commercial platform. Bynario, founded in Milan in 2025 by Alfredo Pesoli, Giancarlo Russo, and Lorenzo Cavallaro, closed a pre-seed round led by 360 Capital with participation from PranaVentures.
The product aims to identify, validate, prioritize, and remediate vulnerabilities through an increasingly autonomous process. It is a crucial distinction: the challenge for companies is not just finding potential flaws, but understanding which ones are genuinely exploitable within their environment and require immediate action.
AI lowers the cost of offensive research
Bynario gained recognition after its researchers deployed frontier models in offensive security operations, identifying vulnerabilities in macOS Screen Sharing technology. The issues were reported to Apple and subsequently patched.
However, the episode highlights a paradox. The very same models that enable defenders to find vulnerabilities faster also lower the cost of research for attackers. If it becomes economically viable to hunt for flaws across a larger volume of software, the scale of issues to manage can outpace the capacity of human teams.
Italy aims to build deep cybersecurity
The round will be used to accelerate platform development, expand the engineering team, and grow in the enterprise market. For the Italian ecosystem, it is an interesting signal: rather than merely building a vertical application on top of external models, Bynario is tackling a deep technical problem where AI, the software supply chain, and security intersect.
