Although more than 120 years have passed since the first scientific work on suicide, humanity is still struggling to effectively reduce the number of suicides and counteract self-destructive behavior. WHO reports from 2010 [51] and 2014 [52] on the prevention of suicide indicate the directions and areas of preventive actions, but they also stipulate that – with a few exceptions – they are unable to indicate clearly proven prevention methods. This work, which is based on analysis of the latest literature (mainly from 2015–2019) shows the upcoming changes in the way we look at the possibilities of such prevention. Literature data suggest that the problem of suicide prevention cannot be effectively solved without changing the paradigm and focusing on new technologies. The current paradigm, which is based on so-called causal modeling, does not work when diagnosing the threat of suicide because the problem of suicide is too complex. Hence, it is proposed that machine learning based on large amounts of data (largely from biological material) should be used to create appropriate diagnostic algorithms. In the future, appropriate smartphone applications could be used to guide patients at risk of suicide. The Zero Suicide model described in the literature draws attention to the organizational needs of preventive therapy and the appropriate motivation of people participating in this therapy. The summary of the work contains the idea that – unfortunately, as in many other health issues, but also in the area of suicide prevention – success will depend not so much on the efforts of researchers and clinicians (because they are usually motivated enough), but on the good will and common sense of political decision-makers.