How can I create a function to replace one string column with another, based on a .startswith() conditional?

Question

I have a spreadsheet where a column (naics2digit) does not match the first two characters of the corresponding column, (naics5digit) which is accurate. I want to find a solution where if naics2digit does not match the first two characters of naics5digit, then to replace the naics2digit column with the first two characters of the naics5digit column.

See the bold values below:

industry	naics_2digit	naics_5digit
Accommodation and Food Services	53	53456
Real Estate	72	72543
Other Services	43	72239
Retail Trade	81	81111
Construction	45	45789
Real Estate	23	23435
Real Estate	53	23333

I have tried using np.where() or conditional for loops with .str.startswith() or logical expressions like naics2digit != naics5digit to return rows where you could update the original table values like the following:

industry	naics_2digit	naics_5digit
Real Estate	53	23333

TO

industry	naics_2digit	naics_5digit
Real Estate	23	23333

Any suggestions?