JPMorgan Chase Dialogue with Zhipu AutoClaw: Why Are Agents Exploding, Does Model Quality Still Matter, and How Can It Be Monetized?

“Why is ‘Lobster Farming’ so popular? How will ‘AI Lobsters’ disrupt various industries? And how can they be monetized?”

According to a China securities research report released by J.P. Morgan on March 12, analysts Xu Wentao and Yao Cheng recently discussed with project managers from Zhipu AutoClaw, providing an in-depth analysis of why products like AutoClaw and OpenClaw have become popular, as well as their future application pathways and business monetization logic.

J.P. Morgan believes that “the importance of products like AutoClaw and OpenClaw is not because they suddenly make autonomous AI commercially mature, but because they significantly lower the barrier for non-technical users to experience intelligent workflows.”

For the market, the core impact is: although the adoption of intelligent agents is expected to expand model usage and infrastructure demand, short-term monetization remains in its early stages. Actual deployment will first occur in relatively structured workflows, rather than widespread fully autonomous human replacements.

Lobster Intelligence Agents Explosion: A Product Design Victory, Not a Model Mutation

Is the recent craze around OpenClaw-like products driven by leaps in model capabilities or by product design optimization? The interviewees’ statements give a clear answer. This popularity reflects improvements in product design and usability, not a sudden mutation in model intelligence.

They emphasized three key factors: “Integration with existing communication tools, persistent memory allowing agents to build user profiles over time, and broader system permissions to expand the actual scope of agent work.”

J.P. Morgan points out that this distinction is crucial. The current hype is driven by better productization and accessibility, meaning user engagement can be expanded before achieving true enterprise-level monetization.

Foundation Model Quality: The Core Determining Business Potential

In today’s landscape of countless intelligent agents, will foundational models be commoditized? The clearest point from the discussion is, “The commercial ceiling of intelligent agents still largely depends on the quality of the foundation models.”

Interviewees repeatedly emphasized that “Intelligent agents are essentially containers or mediators; the model remains the core factor in determining whether tasks are completed accurately and consistently, and whether reasoning depth is sufficient to perform well in high-value contexts.”

J.P. Morgan believes this refutes the idea that the model layer will be fully commoditized in the short term. “Better models should translate into better task completion, stronger instruction adherence, more stable long-context performance, and improved handling of open workflows.” Therefore, increasing adoption of intelligent agents remains a significant boon for leading model providers.

Business Monetization: Still Early, Focus on Structured Tasks

Despite the buzz around intelligent agents, the interviewees implied that “in the short term, the intelligent agent market may still be in an exploratory phase with limited monetization.” Current products are mainly helping users discover use cases. To significantly expand in commercial scenarios, “it may still require 6 to 12 months of model improvements, workflow training data, and product iterations.”

J.P. Morgan suggests that this aligns with the current state of enterprise AI. “Coding and technical workflows remain the clearest early monetization paths because tasks are structured, goals are clearer, and execution trajectories are easier to define.” Beyond coding, the lack of standardized ‘trajectory’ data is a key constraint for agents executing multi-step real-world tasks.

Market Deployment: Prioritizing Technical Engineering and Structured Workflows

Which fields will adopt intelligent agents first? The interviewees highlighted three main categories:

• Technical Engineering: “From coding to testing, deployment, operation, configuration, and debugging. This seems to be the most credible category for today’s business.”

• Information and Content Workflows: “Including research, report generation, document processing, office file handling, and internal content creation.”

• Personal Productivity: “Such as email, calendar, and messaging management.” While attractive to consumers, sustained monetization may take longer.

J.P. Morgan advises investors, “Set short-term expectations on technological and structured enterprise tasks rather than making overly aggressive assumptions based on consumer trial use.”

Open Architecture and Moats: Rapidly Replicable Features Are Not Critical

Another key point from the discussion is that the intelligent agent layer may not be a winner-takes-all proprietary model channel. AutoClaw supports multiple model providers, with management explicitly endorsing an open architecture rather than forcing exclusive use of Zhipu models.

J.P. Morgan sees this broadening the potential market and increasing the chance for intelligent agent platforms to become aggregators of model ecosystems. But for model providers, this means “the agent interface itself may not guarantee exclusive downstream value unless the provider also leads in model performance, tool invocation, and workflow integration.”

Regarding moats, management believes that feature comparison is less important because many visible features can be quickly copied.

They argue that true defensive advantage lies in three areas: “Speed of product insights, foundation model quality, and accumulated intelligent agent functionalities (such as browser tools, memory systems, and workflow processing).”

J.P. Morgan agrees, noting that investors should focus on whether providers can “continuously improve task completion rates, reduce friction, and leverage data to enhance agent performance over time.”

Industry Chain Reshaping: Who Benefits, Who Gets Disrupted?

Broader adoption of intelligent agents will benefit multiple parts of the AI stack:

• First, broader adoption should benefit model providers because more autonomous workflows mean higher token consumption and longer usage durations.

• Second, inference infrastructure, cloud, and related computing providers should also benefit, especially as demand continues to outpace supply.

• Third, open APIs and controlled integration via collaboration and workflow platforms can benefit by becoming interfaces that embed agents into daily work routines.

Conversely, companies whose value propositions are “shallow intermediaries or low-threshold information processing” may face risks. For roles or services with limited moats, open information, and relatively automatable workflows, AI could exert pressure.

Additionally, security and regulation are practical constraints for enterprise deployment. Management notes that “prompt injection, permission errors, malicious third-party skills, and software vulnerabilities are real constraints.” This may slow monetization in the short term but will increase the importance of trusted vendors and compliant architectures.

J.P. Morgan maintains a “Buy” rating for Zhipu, with a target price of HKD 800 by December 2026, based on a 30x expected 2030 P/E ratio, projecting a compound annual revenue growth rate of over 100% from 2026 to 2030.

This insightful content is brought to you by Chasing Wind Trading Platform.

For more detailed analysis, real-time insights, and frontline research, please join the 【**Chasing Wind Trading Platform - Annual Membership**】.

Risk Disclaimer and Terms of Use

Market risks are present; invest cautiously. This article does not constitute personal investment advice and does not consider individual user’s specific investment goals, financial situation, or needs. Users should evaluate whether any opinions, viewpoints, or conclusions herein are suitable for their circumstances. Investment is at their own risk.