Model extraction / weight theft
Adversarially stealing a model's weights or replicating its behavior by querying it, turning proprietary model IP into a security and trade-secret exposure.
grounded in: Latest trends theme: 'a lawsuit between two tech giants over AI trade-secret theft, signaling AI has moved from research to high-stakes IP litigation' (durable underlying technique, not the suit itsel
Connected concepts
Adversarial attacks on AI, Agent security, Open-weight LLMs
Explore it live in the knowledge graph →