Microsoft (Oct 16) and NVIDIA released "Risk Management for Open-Weight Models in Enterprise", the first in a planned joint series. It offers a shared risk taxonomy mapped to ISO standards and the NIST AI Risk Management Framework. The core argument is that weight openness alone doesn't set a system's risk; the use case, deployment and safeguards do. With open weights the enterprise, not the model provider, owns more of the safeguards (filtering, monitoring, access control, update cadence) over time. Useful if you're writing internal policy for self-hosted models.