How should the Fermi paradox affect an estimate of humankind’s likelihood and best means of long-term survival? A significant probability that many other civilizations have been in our situation but failed to become spacefaring increases the probability that our optimal existential risk strategies are costly, likely to fail, likely to leave traces if they do fail, and might require talents that mankind has but that other scientifically advanced species lack. The Fermi paradox implies that we should seek scientific data based on astronomical observations not accessible to civilizations that lived in the distant past, and that we should create machines to flood our galaxy with radio signals conditional on our civilization’s collapse. Our ability to use Bayesian updating on the Fermi paradox reduces the chance that aliens exist but are hiding from us because of their desire to not interfere in our development: giving us a false understanding of the fate of intelligent life in the universe would cloud our understanding of existential risks. The paradox also provides clues as to types of trap that might destroy us. The possibility that our universe is fine-tuned not only for life but also for the Fermi paradox magnifies these results.