Most architects easily forget that there is a mental impact on the people that will use our systems. The main influence here is usually called the automation-induced complacency or over-reliance on automation. We as human have the strange behaviour that we will always favour information from an automated system over all other information. An easy and typical example is a wrong link in Wikipedia where we will believe the link even if the linked information is complete rubbish.
The main problem with automation-induced complacency is that of users over-reliance on automation include inexperience in a task (though inexperienced users tend to be most benefited by automated decision support systems), lack of confidence in one’s own abilities, a reflexive trust of the automated system, a lack of readily available alternative information, or as a way of saving time and effort on complex tasks or high workloads. As such systems can often create a lot of unwanted side effects, often down to the point where the benefits of automation are are completely eliminated by their skill and confidence destroying automation effects.
If you are working as a solution architect in one of these areas it is often good to talk through this over-reliance on automation with the key stakeholders. If they agree there is a quite an easy fix. Just introduce seldom errors in the system, best on a very random base. Sometimes even a manual override to introduce errors will do the trick. So with this little fix you will ensure that automation actually increases productivity instead of destroying it.