A fragmented ecosystem of AI agents has emerged across messaging platforms, each targeting specific use cases from family coordination to travel planning to professional task management. Rather than consolidating around a single dominant platform, the market has splintered into specialized competitors—Instinct commanding $10 billion in valuation whilst niche players like Miso focus exclusively on travel, Fambot on household management, and Stanley on content creation. These agents operate through existing messaging infrastructure (iMessage, WhatsApp, Telegram, SMS) rather than requiring separate app downloads, and they share a common architecture: connecting to users' existing tools, maintaining conversational context, and executing multi-step tasks autonomously. The diversity of approaches reflects genuine uncertainty about which use cases will drive adoption, though the capital flowing into the space—with Tab emerging at $300 million valuation and multiple Series A rounds closing—suggests investors believe messaging-native agents represent a fundamental shift in how users interact with AI.
For CX teams, this fragmentation presents both opportunity and complexity. The agents designed for household and family coordination (Fambot, Ohai, Ollie, Orbits) are essentially building consumer-grade versions of what support teams already manage through ticketing systems—task routing, context retention, escalation workflows, and multi-party coordination. The question becomes whether these consumer agents will eventually migrate into enterprise contexts, forcing CX platforms to compete on agentic capabilities rather than traditional ticketing features. More immediately, the emergence of agents with autonomous capabilities—Instinct's dedicated email addresses, Wajo's Fo with its own phone number and payment card, agents that can place calls and execute transactions—raises the governance challenge that How to keep AI agents within their permissions directly addresses. Teams running Zendesk or Salesforce need to consider whether their current permission models and audit trails can accommodate agents operating on behalf of customers, and whether the security posture required for consumer agents (particularly privacy-focused options like Underdog) will become table stakes for enterprise deployments.
The market's emphasis on messaging as the primary interface also signals a strategic shift away from dashboard-centric tools. Agents like Comma, which operates across browser, computer, and files whilst remaining accessible through Signal and Telegram, or Skye's "agentic homescreen" approach, suggest that the next generation of CX tooling may need to meet users where they already communicate rather than forcing them into dedicated interfaces. For support leaders evaluating whether to build agentic capabilities in-house or integrate third-party agents, the critical consideration is whether your team's workflow should adapt to these messaging-first paradigms or whether your customer base will demand integration with whichever agents they've already adopted for personal use.
We created a list of the most notable AI agents that can live in your text messages, from general assistants to agents designed for families, travel, and work.