Organizations deploying AI agents into enterprise workflows lack systematic methods to measure their economic contribution. Traditional evaluation focuses on isolated technical performance, not the bottom-line impact of hybrid human-AI teams. A new paper titled "Agentomics: Economic Foundations for the Valuation, Attribution, and Pricing of AI Agents in Human-AI Workflows" (arXiv, June 2026) proposes a rigorous economic framework to close this gap.
According to the paper by Zhu Quanyan, the Agentomics framework models a workflow as a configuration of heterogeneous agents—both human and artificial—whose collective performance determines gross value, deployment cost, reliability, and expected failure loss. Workflow value is treated as a team-level quantity that may include complementarities, substitution effects, bottlenecks, and nonlinear production; additive stage-level value is only a special case.
The Agentomics Framework
Agentomics formulates AI deployment as a coalition-formation problem. The coalition value is defined as the incremental net surplus generated relative to a benchmark human-only workflow. This approach moves beyond simple cost-benefit analysis by accounting for interaction effects between agents. The paper then applies the Shapley value—a concept from cooperative game theory—to attribute the economic surplus among participating AI agents. This yields a principled connection among valuation, accountability, and market pricing.
The Shapley pricing equilibrium provides a normative benchmark for assessing whether agent prices reflect their expected marginal contribution. In effect, it answers the question: "Is this AI agent priced fairly relative to the value it adds to the team?"
Workflow Valuation and Coalition Formation
| Aspect | Traditional Evaluation | Agentomics Approach |
|---|---|---|
| Focus | Isolated technical metrics (accuracy, speed) | Economic contribution to team output |
| Value model | Simple input-output | Coalition value with complementarities, substitution, bottlenecks |
| Attribution | Not addressed | Shapley value based on marginal contribution |
| Pricing benchmark | Cost-plus or market comparison | Shapley pricing equilibrium |
The framework explicitly models deployment costs, reliability losses, and failure risk. It does not assume that each agent's contribution is additive; instead, it accommodates nonlinear production functions where the whole may be greater or less than the sum of parts.
Security Operations Case Study
The paper includes a security-operations case study to illustrate how the framework accounts for productivity gains, deployment costs, reliability losses, and coalition-level complementarities in hybrid human–AI workflows. While detailed results are not provided in the abstract, the case study demonstrates the practical application of Agentomics to a realistic operational setting, showing how the framework can guide investment and pricing decisions.
"Workflow value is treated as a team-level quantity that may include complementarities, substitution effects, bottlenecks, and nonlinear production; additive stage-level value is only a special case." — Agentomics paper
Implications for Enterprise Buyers
For CTOs and technology procurement leaders, Agentomics offers a mathematically grounded method to evaluate AI agents beyond vendor benchmarks. By framing deployment as coalition formation, it helps identify which agents generate genuine surplus and at what cost. The Shapley pricing equilibrium could serve as a negotiation tool for pricing software-as-a-service AI products, tying fees directly to expected marginal contribution rather than arbitrary list prices.
Supply chain and logistics technology managers, in particular, may find the framework relevant as they integrate AI agents for demand forecasting, warehouse optimization, and route planning—scenarios where human and AI agents must coordinate under uncertainty. The ability to attribute value to each agent while accounting for bottlenecks and complementarities aligns with the complex, interdependent nature of global trade operations.
While Agentomics is still a theoretical contribution, its formalism provides a foundation for future software tools that could automate the valuation and pricing of AI agents in enterprise contexts. The paper is available on arXiv under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
