You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Such questions are answerable if the defender has the requisite “capabilities,” each of which provides a specific inference that may be enforced upon on its own, or in combination with inferences from complementary capabilities.
I'm not certain I understand what is meant by capabilities -- is this some combination of signals and the ability to assess them?
The text was updated successfully, but these errors were encountered:
They are intended to be high-level functional requirements: A simple statement of the specific fact that the defender has to assert.
As you point out, the implementation of this includes a signal (what is indicated) and assessment (is it true), however I want to keep us shy of defining the exact signal (boolean? enum? scalar?) and exact mechanism of assessment (e.g. signing + certificate validation) at this stage, if possible.
Some example capabilities may be:
Does a given set of requests come from >N distinct devices?
Is this the physical device that the client reports itself to be?
Is this interaction event coming from a human?
Assuming that enough capabilities resonate across use cases and stakeholders, we should be able to group them into a cardinal set of capabilities for discussion. I anticipate that each capability will attract a robust dialogue involving privacy principles, use case criticality to the user and society, and that this will inform the parameters within which we define sources of truth and specific signals.
I'm not certain I understand what is meant by capabilities -- is this some combination of signals and the ability to assess them?
The text was updated successfully, but these errors were encountered: