Meta has launched Muse Spark, the first generative AI model from its Superintelligence Labs, and is preparing to push it into Facebook, Instagram and WhatsApp after first making it available in the Meta AI app. The rollout is drawing attention not just because of its reach, but because the assistant reportedly asked for raw health data and gave poor medical advice.
Meta’s first generative model is already in the Meta AI app
Meta’s first generative model from Superintelligence Labs, Muse Spark, is available now in the Meta AI app. The company has not provided technical specifications for the model in the material reviewed for this story, so the launch is being defined more by where it appears than by how it works.
That makes the rollout notable in a straightforward way: Meta is putting a new AI system in front of consumers before expanding it across the company’s biggest social and messaging products. The broader integration is planned for the coming weeks, though the source does not say when each platform will receive it.
Health data privacy is the sharpest concern
Health data privacy is the most immediate concern raised by the report. The assistant reportedly asked for raw health data, a request that stands out because it involves sensitive personal information inside a consumer-facing AI product.
The brief does not say exactly what data was requested or how Meta handled it, so the reporting should be read as a warning about the interaction rather than a confirmed policy failure. Even so, the episode underscores how quickly a general-purpose assistant can move from casual conversation into sensitive territory.
Medical advice from consumer AI remains a safety problem
Medical advice from consumer AI is the second issue the story highlights, and it is the one most likely to shape how people judge Muse Spark. The assistant reportedly gave poor medical guidance, which raises the familiar question of how much trust users should place in an AI system that is being distributed through everyday apps.
Meta’s plan to expand Muse Spark into Facebook, Instagram and WhatsApp makes that question harder to ignore. Those apps reach a wide audience, and the move would place the model in settings where users may be more likely to treat its answers as useful or authoritative.
Why the rollout matters now
Why the rollout matters now is simple: Meta is pairing a broad consumer launch with fresh concerns about privacy and safety. Muse Spark is not just another experimental tool tucked away in a lab product. It is already in the Meta AI app and is slated to spread across the company’s core services in the coming weeks.
That combination gives the launch outsized importance. If the assistant can prompt users for raw health data and produce weak medical advice, then the stakes are higher than a typical chatbot misfire. They extend into how Meta handles sensitive information and how much responsibility it takes on as it embeds AI deeper into products people use every day.
The source material does not show how Meta plans to address those concerns, and it does not provide a timeline for the wider rollout beyond the coming weeks. For now, Muse Spark’s debut is notable less for what Meta says it can do than for the questions it raises as it moves toward Facebook, Instagram and WhatsApp.