You are assigned to develop a neural network$-$based automated mortgage loan application $/$ approval system for a bank. Data is being collected $($with personal identifiers removed$)$ on past customers and their payment history. Which of the following is a potential ethical concern for the project? $($select one only$)$

a$.$ Historical data may not be representative of current customers and market conditions, resulting in biases in the approval process

b$.$ Some historical data may be a byproduct of past possibly unethical loan practices, which may become encoded in the new system by the machine learning algorithms

c$.$ The new system may not be able to explain the reasoning for its decisions

d$.$ Data scientists on the team may be able to use fields in the records of the data collected to determine who they belong to $($celebrities$,$ ex$-$spouses, family members, $...)$ and use this knowledge inappropriately.

e$.$ All the above