MONOist Japan presents an analytical reading of the relationship between information technology systems in the manufacturing sector and what is called the “death of SaaS,” focusing on the growing role of AI agents in executing business tasks. The core idea is not that software delivered as a service will disappear entirely, but that the way employees interact with enterprise systems may change radically if an agent becomes capable of performing the work instead of the user.
The article is based on a broader shift in the concept of a business system. For more than 15 years, the discussion centered on why companies need systems that employees cannot do without, even if some of their functions are limited or other tasks are completed using Excel. Enterprise resource planning (ERP) and production management systems represented the clearest examples of centralized systems linked to daily operations.
However, recent developments in generative AI and agents capable of carrying out multiple actions have begun to change this starting point. The article mentions Anthropic’s launch in February 2026 of the “Claude Cowork” tool, which can handle local files on a computer and a web browser and automatically perform office tasks and other work. According to the analysis, the development of this type of agent may reduce the need for employees to interact directly with every function in a business system or with its traditional interface.
What Does the “Death of SaaS” Actually Mean?
The article does not use the term to refer to the end of all Software as a Service offerings. It refers to the possibility that some applications will be used less in their current form once an AI agent takes over interaction with systems and performs the required steps on behalf of the employee. In this model, the interface the user browses may become less important, while controlled access to data and the ability to execute actions become more important than the application’s form itself.
The article identifies two points for understanding this transformation. The first is that some systems are designed around procedures and constraints necessary for the work, and an AI agent cannot fully perform its tasks without human intervention or without complying with specific operational constraints. The second is that systems may change as AI advances, with its capabilities gradually integrated into workflows instead of keeping AI as a tool separate from the core system.
Why Are Manufacturing and ERP Systems Affected?
The article places ERP and production management systems at the center of this discussion because they handle core operations that cannot easily be separated from a company’s data and procedures. It notes that these systems have in many cases remained connected to on-premises operating environments, even as cloud software has spread. Meanwhile, global technology companies such as Salesforce, ServiceNow, Workday, SAP, Oracle, and Microsoft are working to develop AI and agent capabilities within their cloud products.
This difference creates a gap between cloud providers and ERP system providers, and it requires companies planning a new ERP project to consider the system’s AI readiness among the requirements of a request for proposal (RFP). System selection is no longer tied solely to current functionality, but also to its ability to accommodate new models for interacting with data and processes when agents become part of the work environment.
What Changes in Practice?
If the hypothesis discussed in the analysis becomes reality, the center of gravity may shift from the application opened by the employee to an AI layer that interacts with several systems in the background. Instead of entering data and navigating between screens step by step, the user may ask the agent to complete a specific task, while the agent accesses the information and performs the actions permitted by permissions and rules.
However, the article does not present this as a settled outcome. The continued need for human intervention in some tasks means that systems will not become worthless simply because agents emerge. An agent’s success also depends on the data it can access, its understanding of the company’s context, and the constraints governing decisions and actions. Therefore, the real change may be a redistribution of roles among the employee, the AI agent, and the business system, rather than the replacement of SaaS all at once.
certi.news’s Reading
The importance of this proposition for technology managers and manufacturing teams lies in shifting the discussion from “Should we buy SaaS?” to a more practical question: “Which parts will the user still need to perform within the system?” The development described puts pressure on business system providers to prove that their products possess not only data and functions, but also provide a secure foundation capable of integrating with AI agents.
At the same time, the article does not establish that the “death of SaaS” will occur or set a timetable for it. Nor does it present implementation results or figures measuring the impact of Claude Cowork or any other tool on ERP systems. What the text confirms is the direction of the discussion: AI agents threaten the traditional model of interacting with applications, while centralized systems, data, and business rules remain decisive factors in determining the speed and limits of the transformation.