A recent article by Ali ibn Ibrahim Ghabban & Robert Hoyland, discussing the earliest discovered Islamic Inscription, a short footnote (#16) states that the Khashnah inscription should be dated to 52AH and not as 56AH as the edition princeps [1]
They do not elaborate on why this reading (citing Fahmi) is correct. But as the Islamic Awareness website gives us access to a fairly high quality photo of the inscription, we can evaluate this ourselves.
I would argue that the inscription quite clearly reads:
اللهم اغفر لجذيم
بن علي بن هبيرة و
كتب لسنة ست وخمسين
'O Allah, forgive Jaḏīm son of ʕalī son of Hubayrah and (he) wrote (it) [or: it was written] in the year six and fifty'
That is, 56AH, as Sharafeddine originally read it.
The reading Fahmi suggests and Ghabban endorses is ثنتين ṯintayn'two' versus ست sitt 'six'. Without dots these two words end up looking very similar, even in the much more calligraphic font that your computer will undoubtedly render this in:
سٮ
ٮٮٮٮں
One obvious difference though, in the two readings, is that sitt would have 4 'denticles', whereas ṯintayn would have 5. No matter how you look at the inscription, there really only seem to be 4 denticles. One could argue that instead, the form written in the inscription represents ثنتن ṯintān with a (not uncommon) defective ā, but this would imply that the numeral would be in the wrong (nominative) case, which at an early date such as 52 or 56AH would be rather surprising. But even if we accept that unlikely reading, there is still an issue. The final letter n in the Arabic script of this period has a distinct descender, which can be seen in the two attestations of بن bin in the image of the inscription. Wheres a final t, looks identical to a final b, without a descender, for which we can compare كتب kataba/kutiba at the beginning of the third line. That shape obviously looks most similar to the word in question.
I wonder whether the line just under the final b was read by the author as the descender of the n, but even if one does that we are left with the line obviously continuing after it. From the picture, this simply looks like damage. As such, I think the dating reads 56AH, and 52AH seems difficult to defend.
[1] Sharafeddine, A.H. (1977) an-nuqūš al-ʔislāmiyyah bi-darb Zubaydah. Aṭlāl 1.