Roosevelt Elementary teacher Linda Walsh uncovered what administrators are calling the school's first "emotional intelligence cheating scandal" when she noticed that twelve students had submitted nearly identical essays for their weekly "What Makes a Good Friend" assignment.

The students, all eight and nine years old, had been using a child-focused AI called BuddyBot to generate responses for social-emotional learning homework. The AI, marketed to parents as a "conversation starter for developing empathy," had apparently been completing the assignments entirely.

"Every essay mentioned 'active listening' and 'being authentic to your core values,'" Walsh said. "Third-graders don't naturally write about 'fostering meaningful connections' or 'creating safe spaces for vulnerability.' Also, half of them spelled 'authentic' correctly, which was suspicious."

Parent Jennifer Martinez defended her son's use of the technology. "BuddyBot helps him express feelings he didn't know he had," she said. "It's like training wheels for emotional intelligence. Plus, his answers are much more mature than the garbage he usually writes about superheroes."

The district's investigation revealed that children had been prompting the AI with minimal information—"write about being nice to friends"—and submitting the results verbatim. One student's essay included the phrase "leveraging interpersonal synergies," which the teacher noted "sounds like something from a corporate retreat."

Dr. Rebecca Kim, a child psychology researcher at Syracuse University, called the incident "predictably dystopian." She noted that AI-generated responses about friendship often emphasize therapeutic language over authentic childhood experiences. "These kids are learning to talk about emotions like HR consultants," she said.

Roosevelt Elementary has banned AI assistance for social-emotional assignments, though three parents have already filed complaints claiming the policy violates their children's "technological accommodation needs." The district is now developing what it calls "human-only interaction guidelines" for future character education curricula.