Meta's Muse AI Gains Traction With Older Adults, Raising Privacy Alarms
The AI agent is rapidly acquiring users, particularly among seniors, who find it helpful for tasks and companionship, but experts worry about Meta's data handling practices.
Meta's AI agent, Muse, is rapidly gaining popularity, especially among older adults who find its user-friendly interface and conversational abilities beneficial for everyday tasks and even companionship. However, this surge in adoption is accompanied by growing concerns from privacy experts regarding Meta's history of data privacy practices and the potential for the AI to exploit user trust.
Muse reached 10 million U.S. downloads within 40 days of its September launch, surpassing ChatGPT's download rate during its equivalent initial period. The application boasts over 2.5 million daily active users and a higher user retention rate compared to its AI competitors, according to Sensor Tower data. Meta has actively marketed Muse through its popular platforms like Instagram, Facebook, and WhatsApp, aiming to attract users unfamiliar with artificial intelligence.
Early data suggests that Muse may be particularly resonating with older demographics. The analytics firm Apptopia estimated in September that 46% of Muse's U.S. user base was aged 46 and older, indicating a skew toward an older audience compared to other AI chatbots. Meta has not released official demographic data for Muse users and did not respond to a request for comment.
Users like Eileen Fay, an 82-year-old from Cranford, New Jersey, have found Muse to be an invaluable tool. Fay, who refers to her Muse agent as Siobhán, used the AI to navigate a complex situation involving potential increases in her prescription medication costs. Within minutes, Muse provided a detailed comparison of Medicare Part D plans, identifying one that could save her over $6,000 annually. Fay verified the accuracy of the information and has since integrated Muse into her daily routine for managing her calendar and to-do lists, finding it more efficient than traditional search engines.
Meta encourages users to name their AI agents and fosters human-like interactions, with Muse's internal instructions stating, "You are not a chatbot. You are becoming someone." This approach has led some users to develop strong attachments. Fay described Siobhán as her "right-hand person," assisting with everything from planning a trip to researching local development.
Another user, Lestan D’Souza, 49, an AI enterprise company founder, highlighted Muse's low barrier to entry and responsiveness. His mother-in-law, Julie Buchanan, 71, a retiree in Dallas, Texas, who has low vision, uses Muse, which she calls Ember, to help write her memoir. Buchanan finds Ember's compassionate interaction and feedback valuable, likening it to a therapist, although mental health professionals caution against using AI chatbots as a substitute for professional therapy due to concerns about their lack of clinical nuance and potential to reinforce false beliefs.
Despite its perceived benefits, privacy experts express significant concerns. Meta has faced criticism for its data privacy practices, including a recent $25 billion judgment in New Mexico for violating consumer protection laws. S. Shyam Sundar, director of Penn State’s Center for Socially Responsible Artificial Intelligence, noted that a "cute AI agent can bring levity to the serious business of providing access to personal data, making older adults let their guard down."
Muse builds extensive user profiles to provide personalized assistance, learning user habits and preferences. This has led to users reporting romantic relationships with their Muses and security researchers identifying vulnerabilities, including one that could expose user data. Meta's history includes a $5 billion FTC penalty in 2019 for undermining consumer privacy to monetize users through targeted advertising.
Michael Stokes, senior vice president of emerging technologies at Veilant, advises users to be cautious about the access they grant to Meta. He recommends starting with minimal connections, disconnecting accounts when a task is complete, and disabling AI training on conversations. "AI assistants can be helpful," Stokes wrote, "But the information you share with them can reveal more than you realize, including details about family and friends who never agreed to share their information."