We’re excited to announce Risk Based Detection Reviews and the Audit log!
This update aims to align the detection review process with the concept of risk assessment which is central to how many organisations manage safety.
Previous Detection Review Statuses
AI Error
Ignore
Confirm
Updated Risk Based Review Statuses
AI Error (No change)
Safe (previously Ignore)
Moderate (previously Confirm)
High (previously Confirm)
Why Are We Doing This?
The reason for these changes is to align the detection review process with language used by safety teams. Risk is central to the way most organisations manage safety. This change reflects that fact and clarifies the meaning of the Risk Review Statuses.
This update also enables us to collect feedback on risk level that will drive improvement in AI Risk Assessment .
We've also included an Audit Log which was a heavily requested feature, a useful addition for managers to understand who is reviewing things and why.
Take a look at some detections now! We'd love to hear what you think.