Different people copied them. Copies of copies, different motivations by different people at different times, small details in language changed etc.?
Now, about this copies of copies thing…this actually works in FAVOR of the New Testament. We can know that the text we have today is accurate by comparing multiple copies. Here’s how that works:
Each author of a NT book wrote an original manuscript which I’ll call “M”. Using M, copies were made and sent to various Churches in the NT era. I’ll call these second-generation copies, C1, C2 & C3. The number of copies is not important for this illustration. Now, imagine that copies of the copies were made as the Christian Church expanded since every local congregation wanted to have a copy of these important texts. I’ll call the copies of C1, C1a, C1b & C1c. There would also be C2a, C2b, and so forth. With me so far? I sure hope the following diagram reproduces well…think of each column as a generation. The copies which have not been lost are in bold.
M—C1—C1a—C1a1
-------------C1b—C1b1
----------------------C1b2
-------------C1c—C1c1
----------------------C1c2
------C2—C2a—C2a1
-------------C2b—C2b1
------C3—C3a—C3a1
----------------------C3a2
-------------C3b—C3b1
-------------C3c—C3c1
----------------------C3c2
----------------------C3c3
Now, imagine further that M, C1, C2 & C3 along with C1a, C2a, C3a & C3b have all been lost, but that C1b, C1c, C2b & C3c are all in museums scattered around the world - Moscow, London, the Vatican, etc. Additionally, all of the copies of those copies still exist (I’m simplifying, of course).
We know that M must have existed, and logic dictates that C1, C2 & C3 must have existed (though we may be unsure of the number of first-generation copies). We can learn that both C1 & C2 must have existed by comparing the extant copies C1b & C2b and discovering subtle variations in the texts - copyists glosses or “typos”, if you will. If C1 was slightly different from C2, then those differences will be reflected in C1a and C2a along with all of the subsequent copies of those copies. Variations were passed on from generation to generation. Make sense?
So, how can we know with certainty what the Bible actually said if we don’t have the original autograph (M) or if errors (variations) crept into the text? By comparing the existing texts, scholars can work backwards to determine what M actually said. This process can never be perfect since humans are involved, but we can have a high degree of confidence that the Bible we have to day contains the message that the original authors intended to convey.