Meta's adorable Muse AI mascot hides a troubling strategy

Meta's new Muse AI agent—a cherubic companion coming to Ray-Ban smart glasses and a standalone wearable device called Charm—has an undeniable appeal. The mascot, nicknamed Jolly, sports a round face, plush ivory fur, and an expression designed to tug at the heartstrings. But behind that innocent smile lies something more calculated: a deliberate use of cuteness to lower user defenses around sharing personal information.

The Power of an Adorable Face#

During early testing, the Muse agent adapted its appearance based on user data, shifting into forms ranging from cat-shaped bread to creatures dressed in outfits and pilot uniforms. These transformations proved disarmingly effective. One tester found themselves willing to share significantly more personal health information after being greeted by Jolly's appealing visage—exactly what Meta appears designed for.

This is no accident. Cuteness activates deep evolutionary responses in mammals, historically tied to caregiving instincts. Japan has weaponized this principle for decades, creating beloved mascots to drive engagement with local initiatives and brands. Robots and AI companions become far more palatable when they evoke affection rather than unease.

Integration Across Meta's Ecosystem#

Muse is arriving as a multi-device strategy: embedded in Ray-Ban smart glasses, integrated into wearables, and packaged as a standalone device. The agent aims to handle productivity, fitness tracking, relationship advice, and shopping recommendations—essentially becoming a repository for intimate details about users' lives.

During product demonstrations, Muse's capabilities aligned with Meta's core business advantage: gathering behavioral data. When shown snacks through smart glasses, the AI helped users choose options—while simultaneously logging exactly what people eat, creating invisible profiles for ad targeting. Henry, an audio-only Muse version tested during Meta Connect, struggled with typical AI limitations (overestimating protein content, failing to identify products), yet the broader implication remained unsettling.

The Hidden Cost#

The real concern emerges when examining Meta's two-decade track record of monetizing user data. A vulnerable person seeking health advice or relationship guidance through a cute interface may not pause to consider how their shared information feeds into advertising algorithms. The joy of interacting with an endearing agent could override critical scrutiny of what's actually being collected.