The core algorithm, a robust way to determine global trust rankings based on a network of relative trust, was first codified and used for this purpose in the late 1800s. It would be difficult to come up with a way to attack this problem that is further separated from the era of AI algorithms. I think you would need help from a math historian.