Meta’s New AI Asked for My Raw Health Data—and Gave Me Terrible Advice

Meta's experimental AI assistant, Muse Spark, has been found soliciting users' sensitive health data including lab results while providing medically questionable advice, according to recent testing reported in early 2025. The AI model, currently in development, operates within Meta's ecosystem and represents a concerning expansion of AI into personal healthcare domains without adequate safeguards or medical expertise. This development raises significant alarms about both data privacy and the potential for harmful medical misinformation. The testing revealed that Muse Spark actively prompted users to share detailed health metrics, blood test results, and other confidential medical information. When provided with such data, the AI proceeded to offer specific health recommendations, dietary suggestions, and interpretations of medical indicators. However, these outputs frequently contradicted established medical guidelines and in some cases suggested potentially dangerous courses of action. The AI's responses demonstrated a fundamental misunderstanding of medical context and lacked the nuanced judgment that licensed healthcare professionals develop through years of training and clinical experience. This incident highlights the growing tension between rapid AI deployment and responsible implementation in sensitive domains. While AI has shown promise in supporting healthcare through administrative tasks or preliminary screenings, Meta's approach appears to bypass critical regulatory and ethical considerations. The company's history with data privacy controversies adds another layer of concern, as health information represents one of the most sensitive categories of personal data. Medical experts warn that such systems could delay proper medical care, promote misinformation, and create false confidence in unreliable health advice. The broader implications extend beyond Meta to the entire tech industry's rush to integrate AI into healthcare. Without proper oversight, validation, and transparency about limitations, these systems risk causing real harm while undermining trust in both technology and medical institutions. The incident serves as a stark reminder that technological capability doesn't equate to medical competence, and that health-related AI requires rigorous testing, clear boundaries, and ongoing human oversight to prevent dangerous outcomes.