[EE] Carbon Ink PCB Insulation Issues?

I've got a commercial application where an overseas "PCB" manufacturer is
attempting to manufacture low cost carbon ink membranes (very similar to
snap dome switch membranes but repurposed for a niche application that uses
its flexibility) with 2 layers of carbon ink separated by an insulating
layer PCB photo resist.

A curious problem that they keep running into is that their insulating
layers are conductive, but only overtop of the carbon traces! Even after
moving from one layer of photoresist/ink to two layers the test coupons
they produce often measure 1-10kohms between overlapping traces
separated by two layers of insulating ink.

They appear to use standard silkscreen printing methods and standard UV
cure PCB photoresist.

I was wondering if any of the veterans on here (those that are left) might
have any insight.

The only thing I can think of is that carbon dust might be
contaminating/transferring/sticking through the thin layers of insulating
ink/photoresist they deposit on the PCBs.

-- 
Jason White