Instank and Muse break into phone networks: AI agent race moves from screens to direct calls

The competition among smart assistants is no longer limited to text chat windows and message exchange; it has moved directly to traditional telephone networks. In a step that reflects independent agents shifting toward performing real-world tasks on behalf of users, both the Instank platform and Meta, through its assistant Muse, launched an external voice call feature, ending the advantage that rival apps had by using phone calls as their sole distinguishing point against leading platforms.

Moving from text interfaces to phone lines redefines the role of the digital agent in daily transactions.I see this shift as a pivotal step that moves the agent from a mere virtual interlocutor to a fully integrated executive mediator. Instank founder Noah Sheen announced on the X platform the launch of Instank Concierge, an advanced service designed to manage sensitive tasks that require direct intervention, such as calling restaurants that do not offer online reservations, placing users on waitlists and canceling appointments at dental clinics, or resolving technical issues and bills with cable companies. This feature is currently in an early access phase for a group of users, paving the way for broader rollout to all subscribers later.

Instank’s move comes amid a rapid technical expansion the platform has recently experienced, having given its users last week custom email addresses that allow the agent to create and manage accounts autonomously, as well as enabling direct communication and coordination among different users’ agents to complete shared tasks via a trusted network. These actions are backed by substantial investment appetite for the company, which is based in San Francisco, having raised $350 million last month at a valuation of $2.5 billion, and is currently negotiating to raise an additional $1 billion at a valuation of up to $10 billion amid growing investor interest in assistants that can operate across multiple platforms.

On the other hand, Meta continues to pursue this path aggressively through its assistant Muse, which was launched only last week. Ryan Fox from the Muse development team confirmed the expansion of the assistant’s beta version to make external calls with companies and stores in the United States, giving priority to users who requested the feature directly from the assistant, the mechanism the company uses to test new features and gauge audience demand. Data from analytics firm Sensor Tower indicates that Muse outpaced Meta AI in its initial growth, recording over 730 thousand downloads in the United States in its first days compared with 707 thousand downloads the previous app logged in its first five days, and peaked at the second spot in the most downloaded apps lists, igniting competition with other market rivals such as Wajo, Town and Olli.

This shift from screens to phone calls has immediate practical repercussions for the institutional landscape in the Gulf, Egypt and the Arab Levant.In markets such as Egypt and Jordan, where contact centers, support services and process outsourcing are the backbone of the digital business sector, support staff and customer service representatives will no longer receive calls only from human individuals, but will find themselves dealing with artificial agents authorized to negotiate, book, follow up and resolve financial disputes on behalf of their owners. This forces local companies to redesign interactive voice response systems and identity verification standards, and to train staff to interact with synthetic callers for whom conventional human conversation methods do not apply.

For institutions and startups in Saudi Arabia and the United Arab Emirates, the growing reliance on phone calls to complete bookings and services means that APIs alone are no longer sufficient to cover commercial interaction. If you lead a technical team or manage business operations in the region, your next step requires reviewing the readiness of the telephone infrastructure to receive inquiries from automated agents, and investing in low-latency interactive voice interfaces that support local dialects, in anticipation of these services reaching regional markets and shifting consumer behavior from manual search to full voice delegation.