We might not notice a 'mass' extinction

Background There is overwhelming evidence that global change is having widespread, detrimental impacts on biodiversity. Population declines and local disappearances have been recorded with increasing frequency across all taxa, resulting in a steady rise in the number of threatened species. However, the number of documented extinctions remains counterintuitively low (~ 1000 species across all kingdoms) compared to the sense of emergency pervading the scientific community. In isolation, that figure might fuel scepticism about the biodiversity crisis, but when put into context, it reveals that current extinction rates might be comparable to those that occurred during past mass extinction events estimated from the fossil record (≥ 75% extinctions within < 2 million years). Although this is an important clue supporting the claim that we might now be witnessing a new ('sixth') mass extinction, it falls short of definitive proof. The claim bears such high importance that it requires exceptionally solid foundations. However, our main aim was not to ascertain whether current extinction rates qualify as a new mass extinction event in progress. Instead, we examined the intersection of potential future loss scenarios and species discovery rates to address the fundamental question of whether and when we will be able to confirm a mass extinction is under way. Advances Our extrapolations suggest that the timing for a mass extinction to materialise (2,604–34,808 years from now at 75% diversity loss) is consistent with past mass extinctions (e.g., 12,000–108,000 years estimated for the Permian-Triassic extinction to unfold) under modern extinction rates (loss of 0.004%–0.053% of global species richness per year). We identify the minimum necessary conditions in which we could confirm a mass extinction under the full range of assumptions related to total species diversity (ranging from < 1.8 million to 163.2 million animal species) and discovery rates (e.g., ~ 13,110 new animal species described per year as of 2026, with the number growing by ~77 species per year), and the associated timeframe required. We show that there are many realistic future scenarios where we would fail to detect a mass extinction in progress. Outlook Based on available evidence, the rate of global biodiversity loss might already be consistent with the standard definition of a mass extinction. But even if true, current extinction rate estimates (20–8343 times background rates) would not necessarily imply a mass extinction is currently unfolding, because this claim can only be verified a posteriori. Our projections instead indicate that there is a high risk of not recognising a mass extinction as it unfolds — 49% across all parametrisations we explored. Furthermore, the temporal scale required for a mass extinction to materialise is orders of magnitude longer than relevant policy and legislative horizons, a mismatch that might appear to absolve today's society of responsibility. In reality, the opposite is true — underestimating the likelihood of already being on a trajectory toward a mass extinction could have catastrophic consequences for future generations and historical accountability. Future generations will be forced to confront a world they perceive as normal, unaware of how much better off humanity could have been.