I found this pretty interesting <https://www.hacks...
# general
m
If I'm reading the article right, the title is somewhat misleading. The researchers did not actually spot Trojans at all. They spotted their own design changes, which I suppose were not related to Trojans. Trojans only come in to things in that the same method could be used to spot Trojans in principle.
m
I read it that they simulated trojans and were able to detect the gds vs Si mismatch. Real trojans would require an actual adversary.