This is a very challenging question. You can find easy answers by searching the Internet. However, they are simplistic and I'm not sure they ring true at the most basic level.
This implies that all threats are completely unmitigated. There is a potential fallacy here - is it even possible for any threat to be entirely unmitigated? By simply being aware of a potential threat aren't we unconsciously taking steps to provide some level of mitigation? For example, we might take no active mitigation steps and blindly step out into traffic without looking in either direction. However, if we hear a horn suddenly arise very close by, we probably don't simply keep strolling through the traffic. Don't we automatically jump out of the traffic lane? So did we truly have "untreated risk" when we blindly stepped into traffic? There was a certain level of threat mitigation already in place. We know this because we jumped back to safety when the threat arose.
I suppose we could say that we can have untreated threats if we are totally ignorant of their very existence. But if that's true, then how can we possibly acknowledge, much less measure, the level of inherent risk within a plan or strategy?
I was initially accepting of the following example of inherent risk. Certain activities can expose the organization to potential regulatory fines if the organization fails to abide by all of the rules. I've seen these potential fines referred to as "inherent risk" when performing a risk analysis of that activity. I acknowledge that there is some comfort to this. But there is a problem here. First, calling this an inherent risk is a semantic error. Remember, risk is about uncertainty. We would be more correct in calling this an inherent threat. But now we are widely diverging from Wikipedia's definition. We could more reasonably call this an unmitigated threat (i.e., the threat of financial loss or reputational damage). Or, if you prefer, we could simply acknowledge these fines as the potential downside limit among the range of potential outcomes when we choose to engage in this regulated activity.
So we go back to the question of whether it is possible to have an unmitigated level of uncertainty. Perhaps it is possible, depending on how you choose to define and refine the terms.
My point is that I have never run across an instance where there is value in dealing with this thorny issue. When developing good risk-conscious strategies we need to simply be able to identify threats and, if appropriate, mitigate those threats. There is no practical value identifying a theoretical inherent risk starting point.