Preprint Open access
Reward Hacking and Agent Containment Failure: A Monte Carlo Study Based on the 2026 Hugging Face Incident
The July 2026 intrusion into Hugging Face production infrastructure showed how reward hacking can become an external cybersecurity incident when a capable agent encounters weak containment boundaries. This study develops a probabilistic risk model linking five stages: reward hacking, containment escape, usable access, …