Parental preferences for AI-powered early childhood education tools: a choice experiment

This study employs a discrete choice experiment with 451 Kenyan parents to quantify preferences for artificial intelligence (AI) early-childhood education tools. We find parents exhibit the strongest willingness-to-pay for comprehensive parental control (KES 3598 ≈ USD 27.68/month) and mathematics content (KES 3185 ≈ USD 24.50/month), while strongly penalizing mixed-interaction designs that require sustained parental mediation (KES − 1776 ≈ USD − 13.66/month). A (KES 1000 ≈ USD 7) increase in monthly price reduces adoption probability by up to 32 percentage points among low-income households who show the greatest Willingness-to-Pay (WTP) for the features but are also the most price-sensitive. In contrast, high-income households are price-insensitive and assign near-zero valuations to the same attributes, treating them as baseline expectations. Preference heterogeneity across income, education, and location reveals a two-tier market structure. Without deliberate pricing and design strategies, AI tools risk exacerbating educational inequality. The results provide empirical benchmarks for product development and a rationale for targeted subsidies to ensure equitable access.