Researchers are developing ethical AI tools to detect self-harm language in youths’ private Instagram conversations, enabling earlier intervention, improved support systems, and safeguards for privacy.