“AI agents — autonomous systems that perceive, reason, and act on behalf of human principals — are poised to transform digital markets by dramatically reducing transaction costs,” wrote Peyman Shahidi, Gili Rusak, Benjamin Manning, Andrey Fradkin, and John Horton in their recent NBER working paper, “The Coasean Singularity? Demand, Supply, and Market Design with AI Agents.”
The authors argue that the broad adoption of AI agents will have far-reaching economic consequences. While the exact outcomes remain uncertain, the underlying forces are familiar: supply and demand will shape how agents are deployed, and technological change will alter the relative costs of economic activities. To understand these shifts, they turn to a foundational insight from economics — Ronald Coase’s 1937 theory of the firm.
Why do firms exist? This question has long been central to economic theory, most notably to the theories of Ronald Coase, the eminent British economist and recipient of the 1991 Nobel Prize in economics. In his 1937 seminal paper, “The Nature of the Firm,” Coase explained that in principle, firms could rely entirely on open markets to procure goods and services efficiently. In practice, however, markets are not frictionless. Transaction costs — searching for suppliers, negotiating contracts, coordinating work, and managing intellectual property — make purely market-based coordination costly and inefficient. Firms emerged as a way to reduce these costs and organize economic activity more effectively.
A firm will expand as long as it is cheaper to perform additional activities internally than to contract them out in the marketplace. But this expansion has limits. As firms grow, they often become more hierarchical and bureaucratic, which can slow decision-making and reduce adaptability. Successful organizations therefore seek an optimal balance between internal production and external sourcing — the classic “make-or-buy” decision.
As the NBER paper notes, many of the activities that constitute transaction costs — learning prices, negotiating terms, drafting contracts, and monitoring compliance —are precisely the kinds of tasks that AI agents can perform at very low marginal cost. If agents can execute these functions effectively, the traditional boundaries of firms may shift significantly, reshaping both organizational structures and market dynamics.
These transformations will not be limited to existing markets adapting to new technologies. AI agents are also likely to enable entirely new, “agent-first” market designs, expanding the frontier of feasible economic organization.
At the same time, the authors caution that AI agents are not guaranteed to improve market outcomes. Even if it is individually rational for firms and consumers to adopt them, the resulting equilibrium may be suboptimal. Externalities across agents, imperfect alignment with human objectives, and asymmetric information could all introduce new inefficiencies. These challenges point to a rich agenda for market design: translating economic theory into mechanisms that can guide the transition to agent-driven markets while capturing their potential benefits.
What Is Driving the Demand for AI Agents?
Humans will demand AI agents for many of the same reasons they rely on human intermediaries. Delegation becomes attractive when the agent’s time is cheaper, when the agent performs tasks more effectively, or when anonymity is desirable.
In relatively simple applications — such as product search — AI agents can automate time-consuming tasks including search, screening, quoting, negotiation, and scheduling. They can also compare large numbers of options in parallel at a fraction of the cost of human effort.
In more complex settings — such as repairing a home appliance or developing a large software system — agents can lower the costs of both exploration and execution. This reduction in cost may enable projects that would not otherwise have been undertaken at all.
The paper suggests that AI agents are likely to gain traction first in markets where delegation is already common:
- High-stakes transactions (e.g., real estate, job search, investments): agents can perform extensive due diligence at near-zero marginal cost.
- Large counterparty spaces (e.g., dating, freelance hiring, rentals): agents can evaluate thousands of options simultaneously.
- High evaluation effort (e.g., venture funding, college admissions, B2B procurement): agents can analyze vast datasets without relying on heuristics.
- Information asymmetries (e.g., used cars, insurance, legal services): agents can continuously monitor and cross-check information sources.
- Experience asymmetries (e.g., home buying, wedding planning, estate planning): agents can leverage insights from large numbers of prior transactions.
How Might AI Agents Transform Markets?
The implications for agents go beyond firm boundaries — they extend to the very design of markets. As AI agents become widespread, markets will need to evolve — not just through policy changes, but through new technical infrastructures and market mechanisms designed to accommodate agent-based interactions.
Identity and Verification
Over time, a significant share of online activity may originate from AI agents rather than humans. This raises fundamental questions about trust and accountability. Ensuring reliable interactions may require new identity frameworks, including cryptographic verification, digital IDs, or decentralized identity systems.
Two broad approaches are likely to emerge:
- Walled gardens, where users authenticate before interacting — though these remain vulnerable to agent amplification.
- Proof of personhood, which relies on biometric or other methods to uniquely identify individuals but would require widespread adoption and systemic change.
Together, identity, credentials, and reputation systems could enable more sophisticated forms of market interaction, including highly customized pricing and negotiation.
Changes to Digital Platforms
Digital platforms will also need to adapt to a world of pervasive agents.
Agents may act as filters, managing information flows and presenting only content aligned with user preferences and goals. At the same time, they may also act as producers, generating content, initiating transactions, and interacting with other agents at scale.
Because these actions can be performed at very low cost, platforms may face a surge in activity driven by agents operating on behalf of users. In response, platforms are likely to differentiate between:
- Agent-oriented interfaces, optimized for machine-to-machine interaction
- Human-oriented interfaces, designed for direct user engagement
Enabling New Market Designs
Perhaps most importantly, AI agents could enable entirely new forms of market design.
By leveraging AI-derived preferences, markets could implement sophisticated matching mechanisms that require detailed preference data across large numbers of alternatives — something that would be infeasible for humans alone.
In addition, agents may help users better understand their own preferences. By observing behavior across many decisions, an agent could identify patterns — such as a consistent preference for certain features in housing — and surface these insights to the user. In this way, agents may not only act on behalf of users, but also help them make better decisions.
Conclusion
“The capacity of AI agents to dramatically reduce transaction costs as automated intermediaries could unlock new forms of market participation, enable previously infeasible mechanisms, and push allocative efficiency closer to competitive ideals,” the authors conclude. At the same time, the very characteristics that make agents powerful — their speed, scalability, and low marginal cost — could also strain existing market structures.
Ultimately, the impact of AI agents will depend on how they are designed, how markets adapt, and how regulatory frameworks evolve. While the capabilities of AI agents are novel, the forces shaping their impact are not. As in earlier technological transformations, fundamental economic principles will determine outcomes.
The structure of agent-driven markets will be shaped by deliberate choices in system design, platform architecture, and regulation. Whether AI agents lead to more efficient, competitive markets — or to new forms of concentration and strategic manipulation — will depend on how these choices are made. Economists, technologists, and policymakers therefore have a rare opportunity: not just to analyze this transformation, but to help shape it.