Introduction
Recently, with the sharp decline in software stocks,The Death of SaaSThe phrase "[...] is becoming a hot topic. This describes a critical situation where traditional business models that rely on UI are being replaced by AI agents."
However, not all companies will be eliminated. The key point is simply whether or not they can adapt to change. By continuing to read this article, you will:
- This will help you clearly understand the reasons for the decline of SaaS.
- You can learn how AI agents work.
- This book offers three strategic proposals for Japanese companies to succeed.
- The potential of a vertically integrated business model can be seen.
Significant changes in the industrial structure are expected between 2023 and 2027. Now is the time for Japanese business leaders to make decisions.
The Essence of SaaS "Death" and the AI Agent Revolution
What does the "death of SaaS" mean? The essence of traditional SaaS companies is simply a "UI (user interface) that connects humans and systems." However, once AI agents can directly operate systems, the UI, which was created for humans, will become unnecessary.
In other words, SaaS is being reduced back to the database itself.
Important evolutionary stages:
- Soft 1.0Traditional software where people write the code.
- Soft 2.0The era of AI writing code
- Soft 3.0The era in which AI can autonomously perform tasks is currently underway.
It just took time for the market to notice this change. In November 2025, a report published by Austrian developer Leonis clarified the organizational differences between SaaS companies and AI companies.
Four SaaS vulnerability categories:
Leovis Capital's analysis categorizes the crisis levels of SaaS companies into four levels:
- Category 1 (Most Vulnerable)These are UI-based business management software programs originating in the United States. They rely on human interaction with the dashboard.
- 2nd category (relatively safe)Areas close to infrastructure, such as cloud monitoring services. Relatively safe because they possess "data gravity."
- 3rd divisionCompanies on the brink of transitioning to AI-powered, native-based systems. The next three years will be crucial.
- Follower forces (the strongest)A startup that built its business around the premise of using AI agents, without being bound by existing UIs.
The most vulnerable Category 1 are SaaS products that rely on human interaction on a dashboard, like Monday.com. These will be among the first to be replaced by AI.
Three Action Plans for Japanese Companies to Win
This book proposes actions that Japanese companies should take to survive in this period of disruptive transformation.
Firstly: Bold investment in a unique domestic foundation model.
We must not rely on inexpensive overseas models and allow our industries to be hollowed out by giant American tech companies, as has happened in Europe. National independence is essential.
Secondly: Breaking free from perfectionism
Japan has a culture of deducting points for mistakes, which is unacceptable when it comes to AI. In the manufacturing era, perfectionism worked, but now management needs to abandon that and be prepared to implement AI even if it means taking the risk of "hallucination."
Third: Building a vertically integrated AI agent
Leveraging Japan's greatest strengths—on-site capabilities and quality control—the key to success lies in building "vertically integrated AI agents" specialized for specific industries such as finance and insurance. A future where human labor and AI work together in parallel will be the key to victory.
Specific organizational differences:
The talent requirements of traditional SaaS companies and AI-first companies are completely different:
Traditional SaaS companies:
- Cloud Architect
- QA Analyst
- SRE (Systems Reliability Engineering)
- Release Engineer
- Middleware Developer
AI-first companies:
- Product Manager
- Front-end engineer
- UX Designer
- Backend Engineer
- Security Engineer
- ML Infrastructure Engineer
- Data Engineer
- AI Platform Engineer
- AI Security Engineer
AI-first companies require a completely different skill set than traditional UI-based SaaS companies.
Supplementary information and points to note
Peter Steinberger, formerly of OpenAI, released OpenClaw, an open-source AI agent. This AI agent will drastically change the lending operations of major Japanese banks.
For example, AI will be able to autonomously handle the entire process, from creating approval documents to manually entering data into SaaS dashboards, without requiring a human to navigate between different platforms. Humans will only need to provide final approval.
Change is coming at an astonishing speed:
- The automotive industry of the 1960s and 70sWith strong top-down decision-making, we were able to achieve transformation at an astonishing speed.
- Now more than ever, decisions from management, not from the bottom up, are needed: Organizations must have underlying strength.
- On-site capabilities and quality control are Japan's greatest strengths:Making the most of this will be the deciding factor in the outcome.
Even in SaaS companies, areas close to infrastructure, such as cloud architects and system administrators, tend to be relatively stable. However, if the business model itself is UI-dependent, there is a risk that it will be completely replaced by AI agents.
Summary
While the death of SaaS is being talked about, not all software companies are in crisis. Only those that can adapt to change will survive. For Japanese companies to succeed, they should build domestic AI, abandon perfectionism, and focus on vertically integrated AI that leverages on-the-ground capabilities.
This is not merely a business topic, but a fundamental shift in the social structure itself. Just as the automotive industry underwent transformation at an astonishing speed in the 1960s and 70s, Japanese organizations still possess the underlying strength to transform not from the bottom up, but through decisions made at the management level.
SaaS is moving towards mere "database creation," and companies that rely too heavily on UI will be eliminated. The skill sets required by AI-first companies are entirely different. The wave of change will hit them faster than most. To survive this disruptive period of transformation, national independence and a departure from perfectionism are necessary.
The key to success lies in building vertically integrated AI agents specialized for specific industries such as finance and insurance, where human labor and AI work in collaboration. Now is the time for management to make a decisive move.
