Instinct and Meta’s Muse are both rolling out AI agents that can place phone calls to real-world businesses in the United States, with a focus on consumer tasks such as booking reservations, canceling or modifying appointments, and handling billing or customer service issues. Human coverage agrees that Instinct’s new Concierge feature is positioned as a full-service text-based assistant that can call, talk to human agents, and complete transactions, while Meta’s Muse is adding similar call capabilities to reach U.S. businesses, bringing both products into closer competition with earlier entrants that had already advertised agentic calling.

Across sources, there is agreement that these moves are part of a broader shift toward “agentic” AI that not only answers questions but also executes tasks in the physical and commercial world on a user’s behalf. Human reporting emphasizes that these systems sit at the intersection of consumer technology and retail infrastructure, relying on connections to real-time data, payment systems, and business operations, while potentially reshaping how advertising, brand preference, and online marketplaces function. Both perspectives treat these agents as iterative steps toward more autonomous, phone- and commerce-integrated AI, rather than as fully general-purpose replacements for existing platforms.

Areas of disagreement

Readiness and maturity. AI-aligned descriptions typically frame Instinct’s Concierge and Meta’s Muse as near-ready or rapidly maturing tools that significantly close the gap between current assistants and fully agentic phone-based AI, highlighting their ability to navigate calls, reservations, and simple problem resolution. Human coverage, by contrast, stresses that these agents are still early-stage, limited by patchy integrations and error risks, and that they are closer to controlled experiments than to robust, everyday infrastructure.

Impact on retailers and platforms. AI-oriented narratives tend to assume that giving agents phone and transaction powers naturally leads to more efficient shopping and service interactions, implying eventual benefits for both consumers and businesses. Human reporting is more skeptical, arguing that agents could undermine retailers’ control over recommendation, search, and sponsored placement, with particular emphasis on how Amazon’s ad- and search-driven model could be “blown up” if agents pick products and merchants on the consumer’s behalf.

Consumer trust and adoption. AI-focused accounts usually present users as ready or eager to outsource more tasks to agents, suggesting that smoother calling and booking experiences will drive quick adoption once friction is reduced. Human coverage highlights significant hesitation, noting that consumers may not yet trust an AI to spend money, negotiate with businesses, or manage high-stakes phone calls, and that overcoming this trust gap is a central obstacle rather than a minor usability issue.

Trajectory toward general agents. AI-aligned views often treat phone-calling abilities as a straightforward stepping stone toward general-purpose AI agents embedded across devices, implying a relatively linear progress path once key features are in place. Human analyses, however, argue that institutional resistance, regulation, and the messy realities of business processes and sales incentives will make the trajectory non-linear, with many of today’s agentic demos remaining constrained by commercial politics as much as by core model capabilities.

In summary, AI coverage tends to portray phone-enabled agents like Instinct’s Concierge and Meta’s Muse as a major, relatively smooth advance toward autonomous consumer assistants, while Human coverage tends to foreground structural friction, platform economics, and user mistrust as critical factors that could slow or distort how these agents reshape retail and everyday interactions.