Nextdoor
Generative AI Kindness Reminder for flagged comments
Nextdoor's AI rewording of flagged comments was adopted by 26% of users who saw it
Problem
Nextdoor needed to keep harmful content low on a neighborhood platform where most reported posts are reviewed by volunteer moderators.
Approach
Nextdoor added generative AI to its Kindness Reminder in 2023. When a comment was flagged as potentially harmful, the tool suggested a more constructive rewording. Machine learning and volunteer moderators handled review of reported content.
Results
Takeaway
Nextdoor reported the 26% adoption for the AI-revised text only. The 15% drop in violating content covers the older reminder and the AI version together.
Categories
- Industry
- Technology and telecom
- Business function
- Production and quality
- Innovation type
- Product and service innovation
- AI technology
- Generative AIPredictive analytics and ML
- Adoption
- Built in-house
- Goal
- Quality and risk
- Evidence
- Company's own claim
Sources
Source summaries
Nextdoor's 2023 Transparency Report release says 26% of users shown AI-revised text adopted it. It says exposure to the Kindness Reminder and the AI version cut creation of violating content by 15%. It also says volunteer moderators reviewed 90% of reported content.
Full details are in the original sources.