AI TraceTrace Foundation, Inc.
Content ModerationInternal OnlyVerified

Reviewed and published by trentmaziarz, March 22, 2026. Discovered and drafted by our automated research pipeline.

Redfin builds automated content moderation into all of its consumer-facing AI chat tools — including Ask Redfin and Conversational Search — to block questions that would require answering in ways that could violate Fair Housing laws. Regardless of how a question is phrased, the system will not provide answers about neighborhood demographics, race, religion, or other legally protected characteristics.

Details

The moderation layer is implemented partly through Sendbird's auto-moderation tools, which sit within the infrastructure supporting Ask Redfin's retrieval-augmented generation (RAG) architecture built on Amazon Bedrock and Amazon OpenSearch. Redfin's SVP of product described the safeguards as surviving extensive adversarial testing — "a lot of rigors" — before launch. In June 2023, OpenAI temporarily deactivated all real estate chatbot plugins, including Redfin's, after an Inman investigation raised Fair Housing concerns. Redfin submitted a 28-page report to OpenAI detailing the risks; the plugin was later restored with enhanced safeguards. Redfin also separately removed crime data from its platform in 2021 to prevent racial bias from entering its models.

Products affected

Ask RedfinConversational SearchRedfin ChatGPT App

Sources & Evidence

Other practices by Redfin

Have evidence about Redfin's AI practices? Submit a report.

Report a Sighting →