The translators of the NET note for Colossians 1:7 and 4:12,
The Greek word translated “fellow slave” is σύνδουλος (sundoulo"); the σύν- prefix here denotes association. Though δοῦλος is normally translated “servant,” the word does not bear the connotation of a free individual serving another. BDAG notes that “‘servant’ for ‘slave’ is largely confined to Biblical transl. and early American times…in normal usage at the present time the two words are carefully distinguished” (BDAG 260 s.v.). The most accurate translation is “bondservant” (sometimes found in the ASV for δοῦλος), in that it often indicates one who sells himself into slavery to another. But as this is archaic, few today understand its force.
The NET seems to be alone among English translations to use "slave" here, but it seems to have good reasoning, and I've heard others say that it's the better translation of the word.
Would the initial readers have understood Paul to be referring to himself and Epaphras and fellow believers as "slaves" in the Greco-Roman sense of personal property?
Would his use of the Greek word have the same scandalous connotation that it would for modern American readers to see "slave"?
If the NET is correct to use "slave" in 1:7 and 4:12, then why not in 1:25?