The start of 2026 has been particularly harsh for software sector stocks. The market abruptly changed its mood, and in many cases, the volatility had less to do with company fundamentals and more with an all-encompassing question: "What does agentic AI mean for the future of software?" An analysis by William Blair illustrates this clearly: the IGV software ETF fell nearly 15% in January, one of the worst starts to the year in its history. The exact figure is less relevant than what it symbolizes: fear is crushing nuance, and the market is treating "software" as if it were a homogeneous block. But the reality is that software is splitting into two very distinct categories.

The bifurcation of software: interface vs. data

The market is punishing software in general, but software is bifurcating. On one side is interface-centric software, heavily dependent on configuration, where the UI is the product. In a world of autonomous workflows, that value will likely erode. Agents will increasingly bypass screens, execute tasks via APIs, and collapse some categories where "clicking" was the primary mode of work.

On the other side are cloud-native, data-centric platforms built to integrate (API-first). This type of software is designed to be operated by both machines and people; it is built for orchestration, integration, and scale. In the age of agents, that is where value will concentrate and multiply.

In other words: sustainable advantage does not come from menus and buttons. It comes from data architecture, semantics, workflows, and connectivity. Romain Boboe's idea of an "interface tax" helps understand this: the more a product's value depends on a person navigating screens and configuring endless rules, the more friction it carries into an era that rewards autonomy.

The missing point: AI cannot (and should not) replace a well-designed SaaS

A simplistic narrative is gaining traction: "AI will generate the application" and enterprise SaaS will become optional. That is not how critical systems work in the real world. AI is powerful and will change how software is built and used. But in many business domains, trying to replace a well-designed SaaS with "custom" systems generated or managed by agents often increases cost and risk, and slows down the business.

There are three underlying reasons why AI will not be the end of SaaS, and why the strongest SaaS organizations are particularly well-positioned to make AI more valuable.

Higher runtime costs

AI will reduce development costs, even significantly. But it increases runtime costs, often notably. Traditional enterprise software is relatively cheap to operate because it is deterministic computation on CPUs. A user clicking "export report" or running a query is predictable and relatively inexpensive.

Agentic AI changes that economy. Natural language questions and answers, summaries, extraction, reasoning, and multi-step agents consume inference computation, often on GPUs or other accelerators. Each interaction can be orders of magnitude more expensive than a traditional database query. Thus, while "creating software" may become cheaper, operating intelligent software will be more expensive and complex.

And that is exactly where "AI-native" SaaS platforms shine: they become the optimization layer between customers and chaos. They keep deterministic systems for what must be fast, cheap, and correct (transactions, permissions, storage, ledgers) and introduce probabilistic AI only where it creates measurable value (assistance, discovery, synthesis, automation). And they do so by orchestrating hybrid infrastructure so that the customer does not have to become an expert in GPU planning, prompt caching, retrieval pipelines, model routing, security layers, and cost governance.

In other words: in the agentic era, SaaS is not replaced by AI; it becomes the platform that controls the cost of intelligence.

More risk and more premium value of service

The "Service" in SaaS is often undervalued by those who expect AI to replace enterprise solutions. But for publicly traded companies, regulated sectors, governments, and any organization concerned with operational risk, the "Service" is essential. Modern enterprises need critical commitments that can be sustained: availability, security and compliance, data governance, support and operations, consistency over time, and accountability when something fails.

If an agent "generates" an application for you, who do you call at 3 a.m. when it fails? Who answers to the regulator if sensitive data is leaked? Who can explain why the system made a decision? When code becomes cheap, risk becomes expensive, and trust becomes more valuable. Operational accountability, governance, and reliability remain premium value, and they do not disappear in the agentic era. If anything, they become more important. As we saw in the OpenClaw case, the lack of clear accountability in AI agents can lead to complex legal and operational dilemmas.

Less time to market: domain depth is worth more than generated code

The vision of "AI does it all" assumes that writing code is the bulk of the value of a SaaS solution. In mature and critical enterprise systems, that is rarely true. In a system like a warehouse management system (WMS), perhaps 20% of the value is the code itself. The other 80% is applied domain knowledge: edge cases and operational realities that are only learned after seeing them in thousands of production environments.

For example: How do you handle a partial receipt when the shipping notice says 1,000 units, but the physical count is 997 and 14 arrive damaged? What is the subsequent impact on allocation, replenishment, and inventory valuation? How do inventory discrepancies trigger different workflows based on SKU turnover, value thresholds, or time since last full inventory?

The teams that build and operate enterprise SaaS have lived through these scenarios repeatedly. That knowledge is embedded in the product, which is why commercial platforms can work from day one of go-live. A custom development, even accelerated by AI, still needs years to discover what you didn't know you didn't know. In other words: incorporating AI makes modern SaaS faster and more efficient. It does not make it dispensable. As Netlify CTO Dana Lawson points out, writing code is no longer the job; the value lies in domain knowledge and integration.

Rethinking the value of SaaS in the age of agents

When the dust settles, AI will not indiscriminately destroy software value. It will revalue software based on architectural realities, rewarding SaaS providers that deliver the traditional value of enterprise software (reliability, governance, accountability, speed of deployment) reinforced with the promise and power of AI.

AI will not be the end of SaaS. It will be the moment the market separates the winners from the rest. Companies that understand that AI is an enabler, not a substitute, and that invest in platforms with solid data architecture, API-first integration, and robust governance, will lead the next decade. In this context, cybersecurity as a prerequisite becomes even more critical, and the bet of giants like Meta on AI agents shows that the race for enterprise artificial intelligence is just beginning.

Original source: ComputerWorld. Analysis and adaptation by ForgeNEX.