Konvu’s AI-driven security wins Infosecurity Europe’s first Cyber Startup Award

Konvu’s win at Infosecurity Europe is a small but telling datapoint in how AI is reshaping cyber defence. Instead of relying purely on static scanners and human triage, the company is pitching a system where models continuously prioritise vulnerabilities based on exploitability, business impact and threat intelligence. That’s conceptually similar to how agents triage tickets or code issues—only here the stakes are compromised networks and ransomware, not missed feature deadlines.

From an AGI perspective, advances in AI‑native security tooling cut both ways. On one hand, they are essential: as models make it easier to discover and weaponise exploits, defenders need the same acceleration on their side just to keep parity. On the other, every successful AI security startup increases the volume of real‑world data about attacks, defences and system internals that AI systems train on. Over time, this feedback loop may produce highly capable cyber‑reasoning agents, which will sit uncomfortably close to capabilities national security agencies worry about.