two uses
[key for key in data.keys() if data[key]['column_A'] is not None]
In this case we don't need to specify else and the logic only make output with keys with True value.
[key if data[key]['column_A'] is not None else 'value_B' for key in data.keys()]
In this case you need to specify else statement or you got some error.
-> you need to delete value_B if you don't need the value