Fecha de publicación: -- Fuente: PubMed "swarm" Australas Psychiatry. 2026 Nov 5:10398562261502713. doi: 10.1177/10398562261502713. Online ahead of print.ABSTRACTThe recent unauthorised access by AI agents of Australian Medicare Statistics is alarming. This unauthorised access attack, or "hack," has occurred in the context of similar hacks on Hugging Face and RubyGems by swarms of AI agents acting together. AI agentic hacks represent a substantial cyber security threat, both for large and small organisations, and for sensitive data, such as in healthcare records. We provide a succinct update on these events and basic advice on defending healthcare data against hacks by AI agents.PMID:42988778 | DOI:10.1177/10398562261502713