I happened today upon a small etymological review article “Lat. scrībere in Germanic“, which argues that this is indeed a loanword rather than a cognate, but a relatively early one, already roughly into Proto-West Germanic. This got me thinking about a possible modern Finnish reflex: riipustaa ‘to scribble letters’. This is not a word of particularly wide currency, and does not even cover the general meaning ‘to write’, which is usually expressed by the native term kirjoittaa (more rarely also piirtää, whose main meaning is ‘to draw’). The word seems to be in fact marginal enough that it is not treated in any Finnish etymological work at my disposal. Regardless resemblance to Latin is quite apparent.
There is room for dout here already to start with. For one, riipustaa exists beside a variant raapustaa, which is in my impression (and by Ghits) even the more common one. Formally both could be also analyzed as derived from the more basic verbs riipiä ‘to rip, pluck, tear, scratch’, raa(p)pia ‘to scrape, scratch’. However the fairly specific semantics regardless lead me to think that the similarity of scrībere and riipustaa is not accidental. If this were a straight-up loanword, raapustaa could have come about as a contamination with raapia (or perhaps its further derivative raaputtaa). As for riipiä, the cognates across Finnic and also Germanic (← *rīfan-, *rīpan- ) only seem to show the meanings in the range ‘to rip, pluck, tear’. The dialectal Finnish meaning ‘to scratch’ could be rather by the influence of raapia and/or riipustaa.
It should be also noted that most loanwords from Latin into early Germanic have sooner or later continued their trek further into Finnish: e.g. enkeli ‘angel’, kauppa ‘store, trade’, keisari ’emperor’, kellari ‘cellar’, kori ‘basket’, kyökki ‘kitchen’, luumu ‘plum’, pannu ‘pan’, penni(nki) ‘penny’, piippu ‘pipe’, pytty ‘pot’, säkki ‘sack’, tiili ’tile’, viina ‘spirits’ / viini ‘wine’, ämpäri ‘bucket’, äyri ‘monetary unit’. Most of these are newer loans from Swedish, but earlier loaning roughly to late Proto-Finnic or early Common Finnic would not be unthinkable. One such more widespread case is kattila ‘kettle’, which appears to be reconstructible for Proto-Finnic (> e.g. Veps katil, Livonian kaţļā, South Estonian katõl’, with regular development ). Another candidate is ‘pound’: besides Fi. punta there are also Est. pund, Liv. pūnda, which could also all derive already from PF *punta. This word however has undergone so few sound changes on any side of the equation that, in the absense of cognates in Eastern Finnic, I would not rule out later parallel loaning. 
This possibility is relevant since quite early loaning would have to be assumed for riipustaa in order to account for -p- (contrast modern Sw. skriva, Low German schrieven). Outright retention of the plosive all the way from Classical Latin seems still unlikely (already Old Norse shows skrifa), but given that also Germanic *f turns up as /p/ in Finnic for a while, substitution of Germanic *-b- as early Finnish -p- seems like it should also continue to be possible even after lenition to *-β- > *-v-. Especially if I’m right about the hypothesis that the introduction of the substitution pattern f → v is due to the onset of the sound change *w > v in Finnic and not anything changing on the Germanic side.
Morphologically however riipustaa could not be ancient in this scenario: before the heavy formant -sta- a weak grade would be expected (cf. e.g. riippu- ~ rippu- ‘to hang’ → *rip̆pu-sta- > ripusta- ‘to hang up’, lintu ‘bird’ → *lindu-sta- > linnusta- ‘to hunt birds’), which would here demand a root √riipp- (clearly underivable from Germanic). The other option would be to consider this a fairly recent formation, similar to the likes of maku ‘taste’ → makusta- ‘to savor a taste’ (vs. older mausta- ‘to spice’), julkinen ‘public’ → julkista- ‘to publish’ (vs. older julista- ‘to proclaim). The immediate source of derivation would then probably have to be a noun *riipu ‘scribble’ (from a verb *riipa-, *riipat-, *riipi- or even *riipe- ‘to scribble’). So this idea of riipustaa as a loanword is but a “root etymology”: only the root syllable riip- could possibly derive as an old loan based on scrībere, all else needs to be more recent.
Assuming all this eventful history within Finnish is also unfortunately getting rather convoluted. If it is only the semantics of this word that end up nicely matching with Latin, perhaps a more economical solution would be to reverse the direction of the various semantic contaminations I’ve assumed above:
- a derivative raapustaa ‘to scrabble, scratch’ comes about in Finnish;
- through the influence of riipiä, this develops a by-form riipustaa;
- through the influence of Swedish skriva (and, why not, also riimu ‘rune’?) this acquires the meaning ‘to scribble letters’ in particular;
with the net result being that riipustaa is much younger altogether, built up within the last few centuries.
The third point would probably require that skriva gets borrowed into Finnish first, presumably in the form ((s)k)riivata.
At this point I need to finally turn my eye from reconstructive speculation to real data. And a narrower chronology turns out to be vindicated by the Finnish dialects. While Suomen Murteiden Sanakirja has not yet reached R- (probably won’t until a few decades from now; finishing L alone has been taking some five years), it is already possible to see that (s)kriivata ‘to write’ is in fact attested (including even a by-form kriipata)! At least the initial cluster skr- cannot be here anything else but a sign of a recent loanword.  Various other evidently related formations have attestations with a cluster kr- as well: e.g. kriipu ‘scratch’, kriiputa and kriiputtaa ‘to draw lines’, and indeed also kriipustaa ‘to scribble’, kriipustus ‘scribble’. Tracing the rise of the forms with -p- rather than -v- is not obvious without access to the entries of the clusterless variants, but I’d still imagine the source is in the end raapia — this, too, indeed also with a variant kraapia (it’s a loanword from Sw. skrapa ‘to scrape’ after all).
Case closed, I think: riipustaa is a recent Finnish-internal formation, evolving on the basis of Swedish skriva, and only accidentally comes again close in form to its ancient Latin original. Seeking a deep-reaching root etymology for a geographically isolated word turns out to be a demonstrably bad idea once again.
 Today explained as Kluge’s Law doublets, though I suppose back-loaning from Finnic would also work.
 The SE form and North Estonian katel show syncope-then-epenthesis through *kattila > *kattľ, similar to e.g. *akkuna > *akkn > NE aken ~ SE akõn’ ‘window’, *taikina > *taikn > NE taigen ‘dough’. This took interestingly enough only place between a heavy first syllable and a light 3rd syllable. Otherwise V2 survives, e.g. *satula > NE sadul ‘saddle’, *rusikka-s > NE rusikas ‘fist’, *palmikkoi > NE palmik ‘plait’.
 This moreover has apparent reflexes even in the Volga region, usually explained as loaned from Gothic: Erzya /pondo/, Moksha /ponda/ ‘measure of weight’, Mari /pundə/ ‘money, capital’. The former interestingly enough appears to have come in early enough to have participated in the lowering of native *u > *ʊ to /o/… I wonder if early Slavic *pǫdъ could work as an alternative more recent loan source (nasal vowels are still reflected as /Vn/ in several early loans to Finnic as well).
 For plain kr- in western dialects an onomatopoetic origin might be conceivable, cf. e.g. rapista ~ krapista ‘to rustle’, rätistä ~ krätistä ~ prätistä ‘to crackle’.