Maria Ressa’s Nobel acceptance speech contains a passage that has aged uncomfortably well. One cup at a time, she likened the task of content moderation to cleaning up a contaminated river. After five years, the cups are smaller, the river is dirtier, and most of the cleaners have been laid off. However, the cups continue to be at the center of most regulatory discussions.
There is a growing debate in academic and civil society circles about this discrepancy between what the issue is and what governments continue to attempt to address. Prosocial tech design governance is a term that hasn’t quite gained traction yet, but researchers like Lisa Schirch at Notre Dame and co-chairs of the Council on Technology and Social Cohesion have been promoting it. It’s awkward. It may also be the most honest way that the issue has been framed in years.
Once you sit with the basic assertion, it becomes difficult to refute. Technology design is not impartial. It was never the case. These options include the autoplay queue, the little red notification dot, the infinite scroll, and the reaction button that rewards outrage more quickly than nuance. In a meeting, someone made a decision. And no human moderator can ever match the scale at which behavior is shaped by the cumulative weight of those choices. The majority of insiders still refer to their own platforms as mirrors, according to Schirch’s interviews with tech workers. Naturally, you are not drawn to the loudest voice in the room by a mirror. Algorithms do.
This framing may be resisted because it is difficult to regulate. It is possible to flag, count, and remove content. The design is less clear. How can a law pertaining to a feed be written? The USC Neely Center has made an effort, creating a design code that allows users to turn on privacy by default and opt out of engagement-maximizing features. Reasonable, nearly dull provisions. This is likely the reason they could be effective.
The research consistently points in the same direction from unrelated angles, which is remarkable. Stronger prosocial tendencies were associated with greater financial discipline and reduced anxiety during economic stress, according to a recent Delhi-NCR region study involving 357 participants. completely different field. The same underlying intuition—prosocial behavior—produces better results overall when it is fostered rather than suppressed. When tech policy wonks and behavioral finance researchers reach similar conclusions without consulting one another, it’s humble.
However, the political timing is awkward. With accusations of censorship coming from all sides, content moderation has turned into a political punching bag. Sensing the trend, tech companies have drastically reduced their safety and trust teams over the past two years. Teams that had been assembled after 2016 were quietly trimmed by Meta, X, and even YouTube. The backlash coincided with the sophistication of the disinformation industry. Poor timing, or perhaps unavoidable timing. It’s difficult to say.
Theoretically, prosocial design provides an escape from the censorship trap. If the platform isn’t designed to magnify the worst speech, you don’t have to decide what constitutes acceptable speech. Prior to resharing, add friction. Alert users when they are about to interact with subpar content. Give thoughtful answers a reward other than dopamine. These are minor issues. They won’t be able to solve every problem. A few of them could be somewhat fixed.
There is a subtle sense that the industry is aware that the current model is unsustainable as this debate develops. Critics did not create the whack-a-mole metaphor; rather, it originated internally. It is genuinely unclear if regulators will continue to treat the symptoms or address the design thread before the next election cycle. There is the research. The more difficult aspect is the political will.
