It sounds like a false proposition at first glance—and it actually is. After all, an Agent has to run on some kind of system, and that system will most likely be a SaaS platform.

Yet behind this question lies a very real anxiety: When AI can automatically generate applications and execute tasks, is there still a need for traditional SaaS?
The answer is clear: AI will not spell the end of SaaS; on the contrary, it is reshaping the industry—not as a destroyer, but as a powerful filter and accelerator.
On one hand, it weeds out SaaS providers that rely on data arbitrage, offer superficial functionalities, and lack customer trust. On the other hand, it accelerates the growth of players who possess deep industry expertise, command full control over workflows, and integrate AI into their core value propositions, propelling them toward a new stage of higher barriers to entry and stronger profitability.
In this transformation, what determines survival or demise is never AI itself, but rather which category your SaaS business falls into. Or to put it another way: whether your SaaS business can pass the dual test of AI.
01 Misread Threat: Which Types of SaaS Are Truly Impacted by AI?
The argument that "Agents will kill SaaS" is rampant, but it is merely a sensational claim made by outsiders.
The reality is that AI has vastly different impacts on different types of SaaS enterprises. We can categorize them into three groups, each with a distinctly different fate:
Data and content-focused SaaS: The most vulnerable, bearing the brunt of the impact
Workflow-based SaaS: Possess moats but still need to evolve
Vertical SaaS: Boast strong moats, with AI amplifying their advantages
Understanding this stratification is a prerequisite for SaaS companies to formulate survival and growth strategies.
02 Category 1: Data and Content-Focused SaaS—Being Filtered Out by AI
This type of SaaS centers its core value on providing aggregated, cleansed, or exclusively licensed data or content—such as market intelligence, news digests, and financial data. They are currently facing a structural crisis.
Why are these SaaS providers so vulnerable?
Public data will eventually be absorbed by LLMs: Large language models are trained on information across the entire internet, rendering the value of "exclusive compilation" obsolete in the blink of an eye.
Synthetic data replacing real data: AI can now generate high-fidelity training data, weakening the scarcity of real data across fields from autonomous driving to medical imaging.
So-called data fences are mostly ineffective: It is extremely difficult to prevent data from flowing into foundational models.
This is not to say that these SaaS companies have no choice but to resign themselves to extinction. There are ways to avoid this fate or prolong their survival.
For example, they can build real-time, dynamic, and non-replicable data loops; or deeply embed data into their proprietary workflows to make it inseparable—turning data into an integral part of the SaaS process rather than a standalone subscription product.
03 Category 2: Workflow-Based SaaS—May Be Accelerated by AI but Requires Proactive Evolution
Cross-industry workflow automation SaaS providers like Asana, Monday.com, and ServiceNow often face the question: "If AI can automatically generate applications, why should we pay for your services?"
Yet this concern overlooks the essence of enterprise-grade SaaS: customers are not buying functionalities or so-called efficiency—they are buying reliable business guarantees.
Why do these SaaS companies still have strong moats? Because:
Customer businesses cannot tolerate "hallucinations": A single error in approval or work order assignment could lead to significant losses.
Compliance and security thresholds are extremely high: Industries such as legal and finance require audit trails, permission isolation, and customer information protection.
Customers need comprehensive solutions, not a patchwork of AI modules: They are unwilling to integrate multiple AI tools; instead, they want a single SaaS product with a clear responsible party and continuous iteration.
This does not mean that these SaaS companies are inherently immune to AI disruption. Instead, they need to identify the right directions for evolution. For instance:
Restructure pricing models: When AI reduces customer labor needs—say, cutting customer service staff by 75%—your revenue will collapse if you continue charging per seat. Therefore, you must restructure your pricing model, shifting toward higher-value usage-based or outcome-based pricing, or even implementing direct price hikes (we have recently seen many SaaS companies raise prices, some doubling their rates).
Proactively integrate AI: Do not wait for native Agent competitors to mature; embed generative capabilities into existing workflows now.
Build a "trust interface" for UI: Many current AI-built UIs fail to clearly distinguish between AI recommendations and system facts. A trust interface can help customers feel in control, becoming the core carrier of a trustworthy AI experience.
The current baseless claim that "Agents will kill SaaS" is mainly an illusion created by VC-driven "AI bubble competitors": they may burn hundreds of millions of dollars to build an impressive demo, which should not be taken at face value.
In reality, the true winner in the rivalry between Agents and SaaS will be the one that can consistently deliver reliable, trustworthy, and auditable intelligent SaaS solutions.
04 Category 3: Vertical SaaS—The Most Likely Winners Accelerated by AI
Vertical SaaS (VSaaS) providers serving specific sectors such as catering, dental clinics, and construction project management, precisely because of their deep focus on a single industry, possess the strongest resilience against AI disruption.
Why do they hold such an advantage?
Control over business critical points: Customers rely on you—not AI—for their core operations. No one would dare let an LLM arbitrarily modify a menu, schedule a surgery, or generate a tax return.
Deep barriers from professional domain knowledge: Industry terminology, rule engines, and years of transaction data are neither publicly available nor easily understandable by general-purpose models.
High switching costs: A SaaS platform may be integrated with hardware, payment systems, and business processes. The risks of replacing a proven SaaS solution far outweigh the benefits.
Because "it works, and no one dares to change it," today’s VSaaS providers inherit this "critical point" advantage and amplify it through cloud-native architectures.
05 How Can SaaS Companies Maximize the Accelerating Effect of AI?
Avoiding the impact of AI is just a defensive strategy; leveraging AI to accelerate your SaaS business is the right way to go with the flow.
Practical actionable steps include:
Build vertical-specific LLMs: Fine-tune open-source models using proprietary high-quality, structured, and years of accumulated transaction data to achieve far greater accuracy and business compliance than general-purpose models.
Defensive UI innovation: Present AI functionalities as "expert assistants," separating uncertainties from confirmed facts. For example: "It is recommended that the compensation amount be 1,800 yuan, based on the average payout for similar cases in the past year."
Defend against AI capability attacks: In certain business scenarios, AI does possess superior capabilities that can significantly impact specific SaaS sectors. For example, credit assessment and big data analytics may erode some of your product functionalities.
The ultimate advantage of VSaaS lies in this: you understand not only SaaS products but also the very essence of the industry. AI will not replace you; instead, it will enable you to deliver deeper, smarter, and more irreplaceable services.
Final Thoughts
AI is not the terminator of SaaS, but a powerful accelerator and filter.
It accelerates the growth of SaaS companies that integrate industry knowledge, trustworthy AI, and deep customer relationships. Meanwhile, it filters out SaaS providers that rely on data subscriptions, offer superficial features, or are overly generalized.
SaaS companies clinging to outdated models will ultimately be eliminated in the AI era. In contrast, those that dare to deeply embed AI into vertical scenarios, restructure product logic, and rebuild trust mechanisms will be propelled to new peaks of growth.
In this AI transformation, there is no "safe zone"—only "action takers." Whether you are accelerated or filtered out depends not on technology, but on making the right choices.

It sounds like a false proposition at first glance—and it actually is. After all, an Agent has to run on some kind of system, and that system will most likely be a SaaS platform.

Yet behind this question lies a very real anxiety: When AI can automatically generate applications and execute tasks, is there still a need for traditional SaaS?
The answer is clear: AI will not spell the end of SaaS; on the contrary, it is reshaping the industry—not as a destroyer, but as a powerful filter and accelerator.
On one hand, it weeds out SaaS providers that rely on data arbitrage, offer superficial functionalities, and lack customer trust. On the other hand, it accelerates the growth of players who possess deep industry expertise, command full control over workflows, and integrate AI into their core value propositions, propelling them toward a new stage of higher barriers to entry and stronger profitability.
In this transformation, what determines survival or demise is never AI itself, but rather which category your SaaS business falls into. Or to put it another way: whether your SaaS business can pass the dual test of AI.
01 Misread Threat: Which Types of SaaS Are Truly Impacted by AI?
The argument that "Agents will kill SaaS" is rampant, but it is merely a sensational claim made by outsiders.
The reality is that AI has vastly different impacts on different types of SaaS enterprises. We can categorize them into three groups, each with a distinctly different fate:
Data and content-focused SaaS: The most vulnerable, bearing the brunt of the impact
Workflow-based SaaS: Possess moats but still need to evolve
Vertical SaaS: Boast strong moats, with AI amplifying their advantages
Understanding this stratification is a prerequisite for SaaS companies to formulate survival and growth strategies.
02 Category 1: Data and Content-Focused SaaS—Being Filtered Out by AI
This type of SaaS centers its core value on providing aggregated, cleansed, or exclusively licensed data or content—such as market intelligence, news digests, and financial data. They are currently facing a structural crisis.
Why are these SaaS providers so vulnerable?
Public data will eventually be absorbed by LLMs: Large language models are trained on information across the entire internet, rendering the value of "exclusive compilation" obsolete in the blink of an eye.
Synthetic data replacing real data: AI can now generate high-fidelity training data, weakening the scarcity of real data across fields from autonomous driving to medical imaging.
So-called data fences are mostly ineffective: It is extremely difficult to prevent data from flowing into foundational models.
This is not to say that these SaaS companies have no choice but to resign themselves to extinction. There are ways to avoid this fate or prolong their survival.
For example, they can build real-time, dynamic, and non-replicable data loops; or deeply embed data into their proprietary workflows to make it inseparable—turning data into an integral part of the SaaS process rather than a standalone subscription product.
03 Category 2: Workflow-Based SaaS—May Be Accelerated by AI but Requires Proactive Evolution
Cross-industry workflow automation SaaS providers like Asana, Monday.com, and ServiceNow often face the question: "If AI can automatically generate applications, why should we pay for your services?"
Yet this concern overlooks the essence of enterprise-grade SaaS: customers are not buying functionalities or so-called efficiency—they are buying reliable business guarantees.
Why do these SaaS companies still have strong moats? Because:
Customer businesses cannot tolerate "hallucinations": A single error in approval or work order assignment could lead to significant losses.
Compliance and security thresholds are extremely high: Industries such as legal and finance require audit trails, permission isolation, and customer information protection.
Customers need comprehensive solutions, not a patchwork of AI modules: They are unwilling to integrate multiple AI tools; instead, they want a single SaaS product with a clear responsible party and continuous iteration.
This does not mean that these SaaS companies are inherently immune to AI disruption. Instead, they need to identify the right directions for evolution. For instance:
Restructure pricing models: When AI reduces customer labor needs—say, cutting customer service staff by 75%—your revenue will collapse if you continue charging per seat. Therefore, you must restructure your pricing model, shifting toward higher-value usage-based or outcome-based pricing, or even implementing direct price hikes (we have recently seen many SaaS companies raise prices, some doubling their rates).
Proactively integrate AI: Do not wait for native Agent competitors to mature; embed generative capabilities into existing workflows now.
Build a "trust interface" for UI: Many current AI-built UIs fail to clearly distinguish between AI recommendations and system facts. A trust interface can help customers feel in control, becoming the core carrier of a trustworthy AI experience.
The current baseless claim that "Agents will kill SaaS" is mainly an illusion created by VC-driven "AI bubble competitors": they may burn hundreds of millions of dollars to build an impressive demo, which should not be taken at face value.
In reality, the true winner in the rivalry between Agents and SaaS will be the one that can consistently deliver reliable, trustworthy, and auditable intelligent SaaS solutions.
04 Category 3: Vertical SaaS—The Most Likely Winners Accelerated by AI
Vertical SaaS (VSaaS) providers serving specific sectors such as catering, dental clinics, and construction project management, precisely because of their deep focus on a single industry, possess the strongest resilience against AI disruption.
Why do they hold such an advantage?
Control over business critical points: Customers rely on you—not AI—for their core operations. No one would dare let an LLM arbitrarily modify a menu, schedule a surgery, or generate a tax return.
Deep barriers from professional domain knowledge: Industry terminology, rule engines, and years of transaction data are neither publicly available nor easily understandable by general-purpose models.
High switching costs: A SaaS platform may be integrated with hardware, payment systems, and business processes. The risks of replacing a proven SaaS solution far outweigh the benefits.
Because "it works, and no one dares to change it," today’s VSaaS providers inherit this "critical point" advantage and amplify it through cloud-native architectures.
05 How Can SaaS Companies Maximize the Accelerating Effect of AI?
Avoiding the impact of AI is just a defensive strategy; leveraging AI to accelerate your SaaS business is the right way to go with the flow.
Practical actionable steps include:
Build vertical-specific LLMs: Fine-tune open-source models using proprietary high-quality, structured, and years of accumulated transaction data to achieve far greater accuracy and business compliance than general-purpose models.
Defensive UI innovation: Present AI functionalities as "expert assistants," separating uncertainties from confirmed facts. For example: "It is recommended that the compensation amount be 1,800 yuan, based on the average payout for similar cases in the past year."
Defend against AI capability attacks: In certain business scenarios, AI does possess superior capabilities that can significantly impact specific SaaS sectors. For example, credit assessment and big data analytics may erode some of your product functionalities.
The ultimate advantage of VSaaS lies in this: you understand not only SaaS products but also the very essence of the industry. AI will not replace you; instead, it will enable you to deliver deeper, smarter, and more irreplaceable services.
Final Thoughts
AI is not the terminator of SaaS, but a powerful accelerator and filter.
It accelerates the growth of SaaS companies that integrate industry knowledge, trustworthy AI, and deep customer relationships. Meanwhile, it filters out SaaS providers that rely on data subscriptions, offer superficial features, or are overly generalized.
SaaS companies clinging to outdated models will ultimately be eliminated in the AI era. In contrast, those that dare to deeply embed AI into vertical scenarios, restructure product logic, and rebuild trust mechanisms will be propelled to new peaks of growth.
In this AI transformation, there is no "safe zone"—only "action takers." Whether you are accelerated or filtered out depends not on technology, but on making the right choices.