You can try to leave a general hedge for unknowns, like never giving anything >99% long term confidence. But actually pricing black swan problems accurately should be impossible, pretty much by definition.
In case of mortgage bonds, the bill payers were bond creators who wanted to get the loans off their books as quickly as possible. The buyers should really know better but they too were largely managers whose salary was paid on % of assets basis with no real downside impact on their own wealth.
Note: making wrong assumptions (explicitly stated) about the future is not fraud. I.e., "I think housing will never go down" is not fraud.