The software-as-a-service (SaaS) industry has been bracing for a so-called 'apocalypse' driven by the rise of artificial intelligence. Analysts warn that AI agents could bypass traditional enterprise software interfaces, directly handling tasks across multiple systems and reducing the need for users to interact with familiar dashboards. Gartner projects that up to $234 billion in application spending could be exposed to this 'agentic arbitrage' between now and 2030, with AI interactions accounting for roughly 20% of enterprise SaaS spending by the end of the decade. Yet, executives from leading vendors like Workday, Freshworks, and Snowflake push back against the notion of an imminent extinction event, arguing that the threat of disintermediation is real but overblown, and that their firms are well positioned to evolve rather than be replaced.

Understanding the Disintermediation Threat

Disintermediation in the context of AI means that an employee might ask a natural-language query to an AI agent — such as Claude, ChatGPT, or Perplexity — and the agent would automatically pull data from various backend systems, compile results, and even execute actions without the user ever opening a traditional SaaS application. This 'ephemeral application' becomes the new work surface. For software vendors that rely on users clicking through their interfaces, this is an existential challenge. However, as Shannon Kalvar, research director at IDC, points out, 'The SaaS apocalypse is overrated, but disintermediation is correct.' The key for vendors is to identify which of their capabilities are truly durable — the unique logic, mathematical models, and trusted data that AI models cannot easily replicate.

Gartner&39;s framing is less about destruction and more about metamorphosis. The analyst firm suggests that SaaS will not disappear but will emerge in a transformed state, with winners defined by how they respond to the threat. Vendors that focus on providing foundational, hard-to-duplicate services — like governance, security, and specialized workflow logic — will remain indispensable, even if the interface through which users access them changes dramatically.

Workday: Trust and the 'Front Door to Work'

Clare Hickie, CTO for EMEA at Workday, articulated her company&39;s strategy during a recent innovation media event at Workday&39;s EMEA headquarters in Dublin. She emphasized that digital leaders are often hesitant to adopt AI due to concerns about trust, privacy, and security. Workday&39;s durable capability, she argued, is its commitment to designing these principles into every product. 'Privacy by design and security frameworks are built into absolutely everything that we do,' Hickie said. This is especially critical for Workday&39;s client base, which includes many risk-averse organizations such as large enterprises and public sector entities.

Workday is building what executives call the 'front door to work' — a central platform where employees log in and use agentic services to ask natural-language questions. For example, a manager could ask about payroll variations and receive a personalized answer sourced from enterprise data. This is made possible by Workday&39;s next-generation service, Sana, which integrates agentic AI across HR, finance, and planning workflows. Hickie believes that as the risk of disintermediation grows, Workday&39;s role as a trusted platform for cross-business workflows will become even more valuable. The company is not merely providing an application; it is offering a system of record that ensures data integrity, regulatory compliance, and auditability — elements that AI agents alone cannot guarantee.

Freshworks: Agile Open Platforms with AI at the Core

At Freshworks&39; Refresh 2026 event in New York City, CTO Murali Swaminathan echoed similar themes. He described Freshworks&39; aim to create an agile, open platform for customer support, with its Freddy AI agentic technology embedded at the core. Swaminathan acknowledged that AI giants like OpenAI and Anthropic are moving fast and have ambitions to capture market share, but he questioned whether they would want to build the underlying system of record required for support management. 'These AI specialists don&39;t want to build everything. They want to be that layer where you engage, but the underlying layer will still be systems like ours,' he said.

He explained that building a support application from scratch using an AI platform is not trivial. Developers would need to code workflows, service-level agreements (SLAs), business rules, and other intricate processes that Freshworks already handles efficiently. Swaminathan also emphasized that the data driving customer interactions will remain within Freshworks&39; system. While AI models can understand requests and generate responses, they cannot replace the operational backbone that governs ticket routing, escalation, and analytics. Therefore, Freshworks&39; durable capability lies in its deep domain expertise and the platform&39;s ability to orchestrate complex customer service operations.

Snowflake: Data as the Ultimate Moat

Benoît Dageville, co-founder and president of Snowflake, shared a similar confidence during the company&39;s Summit 2026 event in San Francisco. He reframed the disintermediation risk by drawing lessons from history. When cloud computing giants like Amazon emerged, many feared that Snowflake would be crushed. Indeed, Amazon launched its own data warehouse, Redshift, as a direct competitor. But instead of being eliminated, Snowflake built a symbiotic relationship with AWS, running its infrastructure on AWS while allowing AWS customers to use Snowflake services. Dageville argued that AWS recognized that being the leading cloud provider meant offering customers choice, not forcing them into proprietary solutions.

Dageville applies the same logic to AI. 'Yes, potentially Anthropic can be Snowflake, or OpenAI can be a Snowflake competitor. That&39;s possible, and then they can build a platform like Snowflake. But what is their magic advantage?' he asked. Snowflake&39;s magic is its deep specialization in data management, security, and analytics — areas that AI companies do not fully understand and are unlikely to prioritize. 'We have the data, and there is no AI without data,' Dageville said. Snowflake&39;s durable capability is its ability to act as the central repository and processing engine for enterprise data, making it indispensable for any AI-driven initiative. The company continuously listens to customer feedback to refine its platform, ensuring that it remains the foundation upon which AI agents rely for accurate and governed data access.

The Broader SaaS Landscape

Beyond these three companies, the broader SaaS industry is grappling with similar challenges. The term 'SaaS apocalypse' gained traction over the past 18 months as market corrections swept through the sector. Investors feared that AI would commoditize software, reducing the value of traditional subscription models. However, many analysts now argue that the real disruption will be more nuanced. Disintermediation will affect vendors that offer only a thin interface layer over generic functionality. But vendors that provide unique intellectual property, industry-specific workflows, or critical data infrastructure will thrive.

For example, companies specializing in enterprise resource planning (ERP), human capital management (HCM), customer relationship management (CRM), and supply chain management all house transactional data and business rules that are hard to replicate. Their interfaces may change, but their role as systems of record will persist. AI agents need accurate, up-to-date data to perform tasks; without authoritative sources, their outputs become unreliable. This creates a strong incentive for enterprises to maintain contracts with established SaaS providers while layering AI on top.

Moreover, the shift to agentic AI does not happen overnight. Organizations must first digitize processes, ensure data quality, and establish governance frameworks. Many enterprises are still early in their digital transformation journeys, meaning the demand for traditional SaaS applications remains strong. Even as AI capabilities mature, the complexity of integrating them into legacy IT environments provides a buffer for incumbents. Gartner&39;s report highlights that the $234 billion exposure is a potential risk, but it also represents an opportunity for vendors that proactively develop services and platforms supporting cross-domain workflows.

Durable Capabilities Defined

So, what exactly constitutes a durable capability? According to the experts, it includes deep domain expertise, proprietary algorithms, robust security and compliance frameworks, and the ability to manage and govern data across disparate sources. It also involves building ecosystems and partnerships that make a vendor&39;s platform sticky. For Workday, it is trust in handling sensitive HR and financial data. For Freshworks, it is the intricacy of customer service workflows. For Snowflake, it is the mastery of data storage, processing, and sharing.

Another critical factor is the ability to offer a unified platform that connects multiple business functions. As AI agents seek to perform tasks that span departments, they will need seamless access to data from HR, finance, sales, and support. Vendors that can provide a single, integrated system — or a data layer that spans these domains — will become indispensable intermediaries. This is why Workday is investing in its platform to cover not just HR but also finance and planning, and Snowflake is expanding its data sharing and marketplace capabilities.

Additionally, vendors are using AI themselves to enhance their offerings. Rather than fighting the trend, they are embedding AI agents into their own products to improve user productivity. Workday&39;s Sana, Freshworks&39; Freddy AI, and Snowflake&39;s Cortex AI are examples of how vendors are turning the threat into a feature. By providing native AI capabilities, they maintain control over the user experience while still offering the underlying system of record. This hybrid approach may prove to be the most effective survival strategy.

The Path Forward

The narrative of a SaaS apocalypse has been a useful catalyst for introspection within the industry, forcing vendors to evaluate where they add real value. The loudest voices of caution have come from those with vested interests in the doomsday scenario, but the evidence suggests a more gradual transition. Software providers that can answer the question 'What durable capabilities are you providing?' will not only survive but could emerge stronger. Their value proposition shifts from 'here is an application you log into' to 'here is a trusted foundation that powers your AI-driven workflows.'

Workday, Freshworks, and Snowflake represent different segments of the market, yet they share a common conviction: data, governance, and specialized logic are their moats. As the industry evolves, the distinction between a software vendor and an AI platform provider will blur, but the need for reliable, secure, and compliant systems will not disappear. The companies that recognize this and adapt accordingly will be the ones that write the next chapter in enterprise technology.

Source: ZDNET News