Key takeaways
- Meta's Muse reached iPad just one month after iOS launch, accumulating 6.6 million mobile installs.
- New connectors bring productivity tools like Asana and Figma alongside retail partners like Walmart and Sephora.
- Websites' bot-blocking defenses can prevent Muse from completing authorized tasks by mistaking the agent for malicious bots.
- Meta is developing an open standard to help AI agents authenticate themselves to businesses and avoid bot-blocking.
Meta announced on Wednesday that Muse, its personal AI assistant, is now available through a dedicated iPad application. The app reached tablets just one month after its September 8 debut on iOS and Android—a pace that underscores Meta’s commitment to the AI agent market, where dozens of startups and platforms are competing to automate portions of users’ daily digital life through conversational interfaces.
Rapid adoption across mobile platforms
Since launching on iOS and Android in early September, Muse has accumulated 6.6 million installs across both platforms, according to market intelligence estimates from Sensor Tower. The speed of the iPad release marks a departure from historical app development timelines. Instagram required 15 years before launching a native iPad application despite sustained user demand, eventually addressing a gap that had frustrated users throughout that period. Muse’s one-month journey to iPad demonstrates how AI agents operate under different deployment pressures than traditional consumer applications.
The rapid transition may reflect technical factors specific to Muse’s design. As a text-based assistant that relies on conversational interaction rather than image or video rendering, the app requires less optimization work to function across different screen sizes than media-intensive applications demand.
What Muse does
Muse functions as a task-completion agent that mediates between users and their digital accounts. Users authenticate the app to access email systems, calendar services, banking platforms, shopping websites, and other online services, then issue instructions in natural language. The assistant can manage inbox and calendar tasks, pay bills, book restaurant reservations, order groceries, set personal goals like savings or fitness targets, and purchase items from integrated retailers. The system is built around granular permissioning: users add service “connectors” selectively, choosing which parts of their online life they grant to the AI. Someone might connect their email and project management tools while keeping banking and shopping under direct personal control.
The expanding connector network
Productivity and business integration
Muse’s capabilities depend on integrations with third-party services called “connectors.” The iPad launch brought multiple productivity additions. Users can now connect to Asana for project management, Figma for collaborative design work, Canva for graphics creation, Dropbox for file storage, GitHub for code repositories, and QuickBooks for accounting functions. Zoom was added for video conferencing and scheduling support. Klaviyo enables integration with customer communication platforms, Notion provides access to notes and databases, and Granola supports meeting transcription. Meta also included connectors to its own advertising platform, allowing agencies and businesses to manage advertising campaigns through Muse.
Retail and shopping partnerships
Meta has integrated major retailers into Muse’s capabilities beyond workplace productivity tools. Best Buy, Gap, Sephora, Walmart, and Wayfair now serve as shopping connectors, enabling Muse to browse inventory and complete purchases across these platforms on users’ behalf. For retailers, the partnerships provide access to a new customer acquisition channel. For Muse, each retail connector expands the range of tasks the assistant can complete.
The bot-blocking obstacle
As Muse operates across retail websites and service platforms, it encounters a longstanding web security challenge. Many websites deploy automated bot-detection and bot-blocking systems designed to prevent unauthorized access, traffic scraping, and fraudulent activity. These defensive mechanisms cannot currently distinguish between a legitimate AI agent acting on explicit user authorization and malicious automated traffic attempting theft or fraud. When Muse attempts to complete a transaction or retrieve information from a site with active bot protections, the website may refuse the request, preventing the agent from fulfilling an authorized task despite the user’s explicit permission.
An industry-wide solution
The open standard approach
Meta announced Tuesday that it is partnering with industry organizations on an open technical standard designed to solve the bot-blocking problem. The standard would enable personal AI agents to cryptographically identify themselves to websites, proving their legitimacy and authorization. By establishing shared authentication protocols and verification methods, the system would allow businesses to distinguish between Muse making an authorized purchase and actual malicious bots. Similar standards already govern browser identification, email sender authentication, and API verification—shared protocols that allow complex web systems to operate securely at scale.
Meta’s position in the agent market
Muse competes in an expanding market of personal AI agents, each attempting to automate aspects of users’ digital workflows. Meta’s structural advantages in this competition include existing user bases across Instagram and WhatsApp, established relationships with major retailers, integrated access to its advertising technology platform, and engineering resources for rapid product iteration. The one-month timeline from mobile to iPad demonstrates the company’s ability to move quickly when a product receives strategic focus. As the AI agent category develops, distribution scale and user access increasingly determine which platforms achieve market leadership.
Source: TechCrunch
Frequently Asked Questions
How many users has Muse attracted since launch?
Since its September 8 launch on iOS and Android, Muse has accumulated 6.6 million installs across both platforms, according to Sensor Tower estimates.
What connectors did Meta add with the iPad release?
The iPad launch brought connectors for productivity tools like Asana, Figma, GitHub, and QuickBooks, plus retail partners including Best Buy, Gap, Sephora, Walmart, and Wayfair.
What is the bot-blocking problem Muse faces?
Many websites deploy bot-blocking systems to prevent malicious traffic, but these cannot currently distinguish between legitimate AI agents acting on user authorization and actual threats, preventing Muse from completing authorized tasks.
