Skip to main content
BackNon-decomissionability of models with open weights
Home/Risks/Gipiškis2024/Non-decomissionability of models with open weights

Non-decomissionability of models with open weights

Sub-category
Risk Domain

Vulnerabilities that can be exploited in AI systems, software development toolchains, and hardware, resulting in unauthorized access, data and privacy breaches, or system manipulation causing unsafe outputs or behavior.

"If the model parameter weights are released or leaked in a security breach, the model cannot be decommissioned because the developer no longer has control over the publicly available model or its use. This prevents effective management and control of an open-sourced or leaked model. Models with publicly available weights are also easier to reconfigure, enabling misuse [178]."(p. 39)

Other risks from Gipiškis2024 (144)