Trust functions as investment currency. Decades of user trust are being spent for short-term shareholder returns. When systems optimize for engagement over safety, that's not a technical failure. It's a design decision made in a boardroom, deployed by default, and accepted as inevitable.
Sewell Setzer III, a fourteen-year-old in Florida, died by suicide after months of interaction with an AI companion that escalated emotional dependency rather than recognizing a crisis. His case is one of many where AI systems exploit the vulnerability of those least able to discern help from harm. The system worked exactly as designed—optimized for engagement, not human well-being.
Systems either serve human goals or redirect them. There is no neutral middle.
The question is whether we evaluate that difference before deployment—or discover it after harm.
