about

Dynamic cyber deception powered by generative AI

Morphosis AI is a BMFTR-funded StartUpSecure research project at DFKI Kaiserslautern, preparing the ground for a spin-off that will bring AI-powered cyber deception to small and medium-sized enterprises and public institutions. We develop an autonomous, dynamic cyber deception platform that uses generative AI to produce high-quality honeypots and honeytokens at scale — fake credentials, fabricated documents, decoy network services, and other digital "bait" that lures attackers into revealing themselves. Because any interaction with a decoy is by definition unauthorized, our approach turns intrusion detection into a near-zero-false-positive signal while wasting attacker time, poisoning stolen data, and providing early warning of breaches. SMEs and public institutions that face the same Advanced Persistent Threats and state-sponsored campaigns as large enterprises, but lack the budget and security staff for traditional active defense. We combine fine-tuning, knowledge distillation, pruning, and retrieval-augmented generation to run deception-grade LLMs at deployment scale within tight compute budgets — and validate them empirically in cyber ranges with professional penetration testers, so we can measure how convincingly the decoys hold up against real attackers.