Luke's genealogy of Christ (Luke 3:23-38) contains an interesting error. Between Shem and Abraham, he is clearly following the genealogy in Genesis 11:10-26. There is one key exception. At one point, Luke adds a name not present in Genesis 11, namely Cainan. This is Luke 3:34-36:
Why does this occur? One explanation suggests itself: Luke was not using the original Hebrew Bible, but rather a Greek translation of it known as the Septuagint or LXX. And the Septuagint does contain Cainan's name (although see the Apologetics Press response, below).
In other words, by using the Septuagint's corrupt translation of the original genealogy in Genesis, Luke may have allowed an error to creep into into his own genealogy.
I suppose someone could argue that our modern Bibles are incorrect, and the Septuagint translation of Genesis 11 gives the original, uncorrupt version. However, there is not a single Hebrew manuscript that supports the idea that Cainan was in the original.
The Apologetics Press argues that, actually, mention of Cainan in the Septuagint was inserted after Luke's gospel was written, by over-zealous scribes, which I suppose is possible. In that case, we can't say that Luke's genealogy was corrupted by using the LXX as his source. We then have no explanation as to where Luke got the name from - though I suppose divine inspiration could be invoked here.
As an alternative, The Apologetics Press suggests the presence of "Cainan" in Luke is itself a scribal error. Although two important manuscripts of Luke fail to include the name, the rest do, and so all modern Bibles contain it. The NET Bible's notes argue that the absence of "Cainan" in two manuscripts of Luke was probably the result of scribes dropping it to be consistent with the original Hebrew.
This whole debate is a mess. I'll summarise the possible positions as I see them:
A key issue is whether the Septuagint available to Luke actually included Cainan's name or not. If it did, it's obvious where he got the name and why his genealogy differs from the Hebrew version of Genesis. The notes to the NIV translation tell us that the LXX did indeed contain "Cainan", as do the notes to the NET Bible translation. Neither gives any caveats about alternate versions of the Septuagint.
However, there certainly would have been different versions of the Septuagint floating around. Every physical copy of a text was the result of scribes copying previous versions, almost always introducing errors or "correcting" mistakes as they did so. I would need to be an expert in Septuagint manuscripts to decide this issue for myself, so I've decided not to classify this problem.