what are the effect on machine learning regression model if the dataset has two exact same columns

What will be the effect on the Machine learning model if the dataset has two exact same columns(exact 1 correlation). One thing that comes to my mind is that if two columns are exactly the same then it is like we are multiplying the column with 2 so in the regression model it is okay since it will take care of that problem by dividing it by 2. So here is the second question if models can work fine then why model have the dummy variable trap since it has a -1 correlation?