P2: 
Where you look, the notification comes in first. 
Then data is collected according to the description of the letter. 
About this specific report from all sources. 
Then it goes into an AI. 
You have rules, an outcome and a presentation of the outcome. 
In the diagram I have tried to capture where do you want the 'decision subject' to have influence? So that landlord or the person who is the subject of the decision. 
You want to see or be able to correct what data is collected about him. 
So what goes into the system then? You want the rules that govern how that system... 
Yes, AI so those decision trees work and what influences that. 
You might also want to be able to see that. 
And possibly also be able to dispute and you have a solution. 
The outcome, we have come up with, is something you might consider, if there is a decision, to visit you or to follow up on such a report. 
That you also have insight into the data and rules that led to that decision to come by. 
So that you kind of get a visual representation of whatever the official saw. 
Or on which he has decided yes I will visit. 
The question mark is still, that cloud is, how can you ensure that the presentation that the official is watching does not also contain a bias, because such a design can also influence that decision. 
And where really, given that civil servant, he can also register that a pattern is emerging somewhere. 
Throughout the time you get various notifications and see the outcome of the AI, so it [unintelligible] just wants to dispute what's included in the rules and how that's being processed. 
And then there was a small cloud as the civil servant, yes, I will pass the decision, that there is still a bias in the assessment of that civil servant himself. 
So you might want to come up with something for that as well. 
That also in that, that is no longer AI, but is an interpretation of it. 
How can you ensure that that part is also secured?