Meta, Instinct and Factory Are Betting Billions on AI Agents

he next phase of artificial intelligence is moving beyond systems that simply generate answers. Technology companies are increasingly building AI that can act on behalf of users, navigate software, complete tasks and operate across digital services.

That shift is creating a new market for so-called AI agents, and capital is already following it. Meta has entered the consumer race with its Muse assistant, while startups such as Instinct and Factory are attracting billions of dollars in private-market valuations. Together, their moves show how quickly the economics of AI are shifting from model capabilities toward software that can actually perform work.

Meta Takes AI Agents to the Consumer Market

Meta’s launch of Muse on September 8 has given the emerging category a major consumer platform.

Muse is designed to perform tasks including sending emails, booking travel and completing transactions. According to Reuters, the application reached 2.8 million downloads during its first 12 days. Meta has also introduced paid tiers priced at $20 and $100 per month, creating an early attempt to turn consumer AI agents into a direct revenue stream.

The significance for Meta extends beyond another AI product. An agent that can execute tasks could become a new interface between users and the digital services they already use. Instead of opening several applications and completing each step manually, users could increasingly give an instruction to an AI system and allow it to coordinate the process.

Meta also has an unusually large distribution advantage through its global family of social and messaging platforms. That makes the company's entry relevant not only to the consumer AI market but also to startups attempting to establish themselves before the major technology platforms fully occupy the space.

Instinct Is Putting a $10 Billion Price on Personal AI

The startup market is moving even faster.

Instinct, the personal AI assistant developed by Spear Street Technology, is reportedly in discussions to raise approximately $1 billion at a valuation of around $10 billion. Sequoia Capital, Benchmark and Coatue Management have been reported as potential participants or lead investors, although the financing has not been finalized.

The proposed valuation represents a dramatic increase from Instinct’s most recent financing. Less than a month earlier, the company raised $250 million at a valuation of approximately $2.25 billion before the new money, according to PitchBook data cited by The Information.

Instinct has reportedly passed 100,000 users, while its limited availability and growing demand have also created computing-capacity constraints. Users can communicate with the assistant through messaging services and give it access to information needed for tasks such as managing calendars, ordering supplies and booking services.

The investment story is therefore not simply about another chatbot. Investors are placing a substantial valuation on the possibility that an AI agent could become a persistent digital assistant capable of interacting with the services people use every day.

Factory Shows the Enterprise Side of the Agent Economy

The same transformation is taking place inside companies.

Factory, a startup focused on AI agents for software engineering teams, raised $200 million in its latest funding round in September. The transaction increased its valuation from less than $1.67 billion to approximately $5 billion, according to Reuters.

Factory’s model illustrates the enterprise version of the agent economy. Rather than simply helping programmers generate snippets of code, AI agents can participate in broader engineering workflows, potentially handling multiple stages of software development.

For businesses, that distinction matters. The commercial proposition of AI is increasingly moving from productivity assistance toward task execution. The question is no longer only whether AI can help an employee work faster, but how much of a workflow can be delegated to software.

That creates a much larger potential market for AI companies.

The New Competition Is About the Interface to Work

The growing valuations of companies such as Instinct and Factory also reveal a change in how investors are thinking about AI startups.

The first wave of generative AI was dominated by foundation models and chatbots. The next layer is increasingly being built around applications that connect those models to real-world workflows.

This creates opportunities across software development, recruitment, finance, customer service, shopping and enterprise operations. Each sector has processes that involve repetitive digital actions, and AI agents are being designed to operate across those processes rather than simply provide information about them.

At the same time, the infrastructure required to support agents is becoming a market of its own. Identity, authentication, payments, permissions, security and reliable access to external software become significantly more important when an AI system can act without a person manually completing every step.

Stripe, for example, has begun integrating its Link digital wallet with personal AI agents from Meta, xAI and other AI companies, an early indication that payment infrastructure is adapting to a world in which software itself can participate in commerce.

The Economics of AI Are Moving Down the Stack

The emerging agent market could change the relationship between users and software.

Traditional applications depend on users navigating interfaces. AI agents could eventually reverse that relationship: users state an objective, and software determines which applications, services and actions are required to complete it.

That could make the agent itself the primary interface to a much larger part of the digital economy.

For investors, this creates an enormous opportunity but also substantial uncertainty. Instinct’s reported $10 billion valuation is based on a financing process that has not yet closed, while Meta is entering the same market with substantially greater distribution and infrastructure. Factory’s enterprise valuation similarly reflects expectations about the adoption of AI-driven software development.

The technology is also still developing. Reliability, privacy, security, computing costs and the willingness of users to give autonomous systems access to sensitive accounts will determine how quickly agents move from demonstrations to everyday infrastructure.

For now, the direction of capital is clear. Meta is bringing agents to millions of consumers, Instinct is attracting extraordinary private-market attention, and Factory is demonstrating the enterprise opportunity. The competition is increasingly shifting from who can build the smartest AI model to who can turn intelligence into software that actually gets things done.