Meta launches Muse personal AI agent on iOS and Android amid privacy backlash and $17 billion settlement
Meta has launched Muse, a personal AI agent powered by the Muse Spark 1.3 model, now in closed alpha testing on iOS and Android. The app, internally codenamed Hatch, offers free and subscription tiers ($20 or $100 monthly) and can autonomously book appointments, fill forms, and monitor home security. The launch follows Meta’s agreement to pay nearly $17 billion to settle a lawsuit over harm to children, and occurs amid ongoing privacy backlash and Wall Street pressure to show returns on AI investments.
Editorial responsibility
- No named human review is recorded for this page.
- Reports are grouped by semantic similarity and deterministic rules. Language models may assist titles, summaries, translation and cross-source analysis; the page itself is projected from evidence records.
- Current automated evidence projection
Cross-source coverage
Wire timeline
Meta has just released the Muse AI assistant for personal tasks, powered by Muse Spark 1.3, Meta's latest model for longer-horizon agentic work. - It can connect to email, calendars, shopping, payments, and other services, then continue working.
Meta has released Muse, a new AI assistant for personal tasks, powered by its latest Muse Spark 1.3 model designed for longer-horizon agentic work. The assistant can connect to email, calendars, shopping, payments, and other services, and continues operating even after the app is closed. Each user receives a Muse Secure VM with a Sentinel agent that inspects outbound data and actions. Meta is offering Muse free for up to 100 million tokens per week, with subscription plans at $20 and $100 per month for heavier usage. Sensitive actions like purchases or sending emails require user approval, and credentials are stored in Secure Credential Storage, preventing Muse from accessing passwords or payment details. The system includes multiple security layers, including VM isolation, credential separation, and a kernel-enforced gate on network actions. The assistant is launching for U.S. users on web, iOS, Android, and WhatsApp.
Meta launches Muse personal AI agent and publishes safety system deep dive
Meta announced the launch of Muse, a personal AI agent, via a post on X. The agent is designed to get to know users over time, holding significant context about them. Alongside the launch, Meta published a detailed deep dive explaining how safety was built into the system, emphasizing that Muse was built to be secure, safe, and private. The announcement highlights the balance between the agent's usefulness, derived from its contextual awareness, and the need for robust privacy protections. The post includes links to the Muse platform and the full safety deep dive.
Meta launches Muse, a personal AI agent with its own browser and WhatsApp interface
Meta has introduced Muse, a personal AI agent that operates on a persistent, isolated Linux virtual machine. Users can assign tasks to Muse via the Muse app or WhatsApp, and it utilizes connected apps and its own browser to complete them. Unlike other AI agents such as Grok Bot or ChatGPT Work agents, Muse is positioned as a low-barrier entry point for users already within the Meta ecosystem, particularly those with little prior experience using AI. The product page describes it as a tool for automating tasks, making advanced AI capabilities more accessible to a broader audience within Meta's platforms.
Show 5 older updatesHide older updates
Meta launches Muse Agent AI assistant powered by Muse Spark 1.3 on iOS and Android
Meta has announced the launch of Muse Agent, a new AI-powered assistant running on Muse Spark 1.3. The agent is available on both iOS and Android platforms, though it appears to be geo-restricted to certain regions. According to the announcement, Muse Agent operates with its own computer system, including a file system and terminal, enabling it to write code and build tools necessary for tasks. It also has access to a full web browser, allowing it to search websites, navigate pages, fill out forms, and complete transactions such as booking and purchasing. This release marks a significant step in Meta's AI agent capabilities, offering autonomous task execution directly on mobile devices.
Meta launches Muse personal AI agent amid privacy backlash and $17 billion settlement
Meta has introduced its Muse personal AI agent app, available in a free tier or monthly subscriptions of $20 or $100 depending on usage. The app, internally codenamed Hatch, is powered by the Muse Spark family of foundation models and allows users to offload digital tasks such as booking appointments, filling out forms, and monitoring home security feeds. Meta AI chief Alexandr Wang stated the app operates in an isolated environment and never accesses users' passwords or payment details. The launch comes at a fraught time for Meta, which recently agreed to pay nearly $17 billion to settle a lawsuit with state attorneys general over harm to children on its platforms. The company also faces ongoing personal injury and school district lawsuits, rising cybersecurity concerns about AI agents, and public backlash against AI data centers. Meanwhile, Meta is under Wall Street pressure to show returns on its massive AI investments, with CEO Mark Zuckerberg positioning personal agents as the foundation for future products and revenue streams.
Meta introduces Muse, a personal AI agent powered by Muse Spark 1.3
Meta has announced the launch of Muse, a new personal AI agent designed to assist users with various tasks. The agent is powered by Muse Spark 1.3, a proprietary AI model. The announcement was made via a post on X (formerly Twitter) from the AIatMeta account, which also included a link for an inside look at how Muse was built and its capabilities. This release marks a significant step in Meta's development of consumer-facing AI tools, aiming to provide a more personalized and efficient digital assistant experience.
Meta's new AI agent Muse seeks access to emails, calendars, payments, and health data
Meta has introduced a new personal AI agent named Muse, which requests extensive access to users' personal data including emails, calendars, payment information, and health services. This move represents Meta's most significant consumer AI initiative to date and poses a critical test of public trust in the company's data handling practices. The announcement highlights the ongoing tension between advanced AI functionality and user privacy concerns, as Meta seeks to integrate its AI deeply into users' daily lives. The success of Muse will likely depend on whether users are willing to grant such broad permissions to a company with a history of data privacy controversies. The development underscores the growing trend of AI agents requiring comprehensive data access to deliver personalized services, raising questions about data security and user consent in the AI era.
Meta's Muse AI agent enters closed alpha testing with proactive goal-tracking features
A new post on X reveals an early look at the user interface of Meta's upcoming Muse agent, based on an App Store listing. The Muse agent is described as a proactive assistant designed to help users track their goals, schedule, and tasks. According to the post, the Muse agent has transitioned from internal development to closed or limited alpha testing. Additionally, new functionality for invite codes has been added, suggesting a controlled rollout for early testers. The post signals that Meta is advancing its AI assistant capabilities, moving Muse from an internal project to a testable product. The exact release date for a public launch remains unconfirmed, but the addition of invite code functionality indicates a phased testing approach.