AI Control Tower and risk integration backend functioning

[AvantikaM]

Hi
Can anyone please explain the backend functioning of the "assessment responses" in the onboarding of an AI asset that automatically maps the risks accordingly and evaluates the risk score and analysis to consider. Like how we map the responses with the risks/control.  

[yashkamde]

Hello @AvantikaM ,

Assessment responses aren't dynamically matched to risks by AI, the mapping is pre-built into the template (question > Metric Category > risk or control record). Responses are scored per question, rolled up via the Scoring Engine (with optional weighting) to a numeric score, which then maps to a rating band (Low/Medium/High) stored against the risk assessment instance and reflected back on the AI asset record.

Refer this :

Smart Assessment Engine Speed Learning— How to use it across your GRC workflows (with video tutorial... 

If my response helped mark as helpful and accept the solution.

[Tanushree Maiti]

Hi @AvantikaM 

Refer:

https://www.youtube.com/watch?v=PNFaNDQ6TZc&t=29s

https://www.youtube.com/watch?v=sV0OgCbyq5U&t=840

Initiate AI assessment on an AI asset 

AI Risk and Compliance release notes 

[OliverV]

The Impact Assessment (and some of the other out-of-the-box AICT smart assessments) has a bunch of automations that will assign Risks and Controls.

Under the Assessment Workspace, open the AI impact assessment assessment. Under the Automations tab you'll see several published automations. 

The Risk Statements mapping automation uses a 'Map AI Risk Statements to AI System' action to assign existing risk statements to the AI Asset if a question is responded to with a 'yes'.

Similarly there are automations for controls that use the 'Map AI Control objectives to AI System'