Edited by: Meredith Peebles
Earlier this month I spent two days in San Francisco in a room with funders, researchers, and practitioners at the
Fund to Prevent Sexual Assault's inaugural convening. I haven't stopped thinking about it since.
Nearly half of adult women and roughly one in six men in the US will experience contact sexual violence in their lifetime. Most are first victimized before they turn 18. And here's the thing that changed how I think about this issue: more than 70% of childhood sexual victimization is perpetrated by other youth, not adult predators. Usually peers close in age. Usually in ordinary social settings. The strongest predictor of who does the harm isn't some individual pathology but rather if their peer group approves of aggression.
To throw a few more numbers at you, we spend an estimated $5.4 billion a year in this country on the justice system's response to childhood sexual abuse. Conversely, we spend a mere $2 million a year on research in the United States that would tell us how to prevent it. Total philanthropic giving to sexual assault prevention and response, across every nonprofit in the US, came to $282 million in 2022. This comes out to about $28 per victim, against a lifetime cost to survivors of roughly $170,000.
This is a gap that's getting harder to close the longer we wait, because the terrain keeps moving. Teenagers are online constantly now, and a majority of boys report seeing content that pushes a pretty narrow, dominance-coded version of masculinity. They say this content "just started showing up” in their feeds; it was nothing they searched for. AI companion apps are quietly filling a role that used to belong to peer relationships, minus anything resembling consent. Most 13–17-year-olds report using AI, with over 40% of kids who feel lonely reporting using it for practicing conversations or social skills and nearly half using it for discussing feelings or personal problems. Consequently, we have a generation of kids being shaped by tools that weren’t designed with healthy relationships in mind.
Most of what currently exists in sexual violence prevention was built for a different decade: school programs designed before smartphones existed, tested on populations that don't look like the kids actually at risk, measuring whether attitudes shifted rather than whether anyone stopped causing harm. None of that holds up against what we now know about the causes of sexual violence, particularly in the age of AI.
When the evidence base is this thin and the cost of getting it wrong is this high, you want to be testing ideas fast, with real evaluation built in from the start, rather than waiting around for someone else to prove a concept first. An incubator model is a great approach to this kind of greenfield as it generates a significant amount of innovation in a short period of time and identifies the people who are interested in solving this problem.
Based on the Demo Day for the
first cohort of the Fund to Prevent Sexual Assault’s incubator, I’m optimistic in this model for catalyzing the ideas and people who can solve this problem at scale. Six months in, the three teams have already gone from concept to a working prototype in testing, and all three have learned something concrete about their users that they didn't know at the outset.
I'm eager to keep watching this cohort, eventually evaluate what they’ve built, and see what cohort two proposes, all in service of the Fund's vision to cut rape in the US by 70% by 2050 (the same kind of generational drop we've seen with teen pregnancy, smoking, and drunk driving).
