The Evolving Landscape of Software: A Shift from SaaS to Agentic AI
Introduction to Agentic AI and the End of SaaS
A few weeks ago, I ventured into the discussion of what I termed the impending death of software as a service (SaaS). My premise was straightforward: with the rise of agentic AI, enterprises can create customized tools tailored to their specific workflows. This capability allows them to surpass vendor-built platforms that aim to serve a mass market. The reliance on generic software was largely an economic necessity during the cloud era, where scale demanded a degree of standardization.
Validation of a Revolutionary Thought
Since publishing my initial thoughts, I’ve noticed validation emerging from multiple fronts. Analysts, tech media outlets, and investors are echoing similar sentiments: agentic AI is compressing the software layers we have come to know. The implication? SaaS may not just face threats from startups but from powerful AI technologies enabling customized solutions.
The App Economy in Jeopardy
As I continued to ponder this shift, I watched the unveiling of Samsung’s Galaxy S26. Suddenly, the disruption appeared to extend beyond SaaS. If agentic AI can destabilize enterprise software, could the app economy be next?
Consider your typical use case of ordering dinner through various apps. You navigate multiple interfaces for reviews, delivery times, coupons, and order tracking. Each app introduces its own friction; they operate as intermediaries between your intent and the final fulfillment of your meal. None actually prepare the food; they merely facilitate the connection.
Transitioning from Apps to Intent
What distinguishes the Galaxy S26? While marketed as an “agentic AI phone,” the real significance lies in the architectural transformation it embodies. For the first time in a widely accessible smartphone, multiple AI agents are integrated at the operating system (OS) level. These agents can act across apps, orchestrating multi-step workflows on behalf of the user.
Changing the Interaction Paradigm
The most significant shift in interaction is rooted in the concept of intent. Traditionally, mobile interactions have revolved around apps—each action dictated by a specific vendor’s ecosystem. Now, the interaction starts with what the user wants, and the OS agent determines how to achieve that goal. The app executes the task but fades into the background, becoming more of an infrastructure rather than the focal point of user interaction.
The Abstraction Layer of Agentic Systems
At this juncture, the agentic layer operates as an important abstraction. The OS agent continues to call upon apps like Uber, DoorDash, and OpenTable, maintaining deep integration within the transaction process, yet significantly enhancing the user’s experience. Users interact with a singular conversational layer instead of juggling multiple dashboards, making the process more fluid.
We can observe a parallel trend within enterprise environments. Agents are increasingly positioned on top of CRM and ERP systems, allowing actions to occur without demanding a series of human clicks. This shift illustrates that while dashboards serve a purpose, they no longer form the primary touchpoint for interaction.
Questioning the Necessity of the Middle Layer
This scenario raises challenging questions about the necessity of the middle layer. Phone manufacturers control numerous elements that shape the digital experience: GPS for location services, cameras, microphones, and powerful edge processors capable of running AI models locally. They oversee authentication frameworks and payment methods, effectively managing user interactions.
Given these factors, why does the interaction chain need to be: operating system → app → vendor API → merchant system? Why not streamline it from the OS agent directly to the merchant?
A Glimpse into the Future
Imagine a future in which you simply say, “Order my usual sushi.” Your phone’s AI agent directly communicates with the restaurant, confirming your identity, negotiating pickup time, processing payment, and providing a confirmation—all without third-party apps or market interfaces.
Restaurants are already adopting AI phone systems; voice models improve rapidly, and the edge computing power of mobile devices is evolving. Each of these building blocks exists in some form, indicating that the trend is not merely speculative.
Entering the Robot-to-Robot Economy
Taking this thought experiment further, envision a restaurant equipped with its own AI system. Your phone’s agent speaks to the restaurant’s AI agent, confirming details without human intermediaries, effectively eliminating marketplace fees and manual navigation.
The implication is clear: agentic AI can craft custom enterprise solutions and reduce reliance on generalized SaaS. The potential exists for consumer software layers to flatten, especially those focused on mediating transactions.
Observing Historical Patterns
Technological history reveals a sequence of transitions where physical storefronts gave way to websites, then to apps. Each shift has minimized friction and streamlined orchestration. Now, we witness another layer of compression with the emergence of agentic systems designed for a world where human interaction is expressed through simple statements of intent.
While the Galaxy S26 does not signify the immediate end of the app economy, it does suggest that apps may no longer serve as the definitive endpoint for mobile engagement. With advancements in edge processing and the growing capabilities of AI agents, the need for numerous, purpose-built interfaces diminishes.
The Economic Realities of the App Ecosystem
However, an underlying story remains unexamined: the economic implications of diminishing reliance on traditional apps. The app economy has generated considerable revenue, with platforms like Apple and Google not only hosting apps but also extracting value from them. This introduces a contentious dynamic, especially with the long-standing 30% transaction fee that has sparked significant legal battles.
As agentic OS reduce the need for traditional apps, how will this revenue model adapt? Will companies like Apple and Google pivot to capture commissions at the agent level? What new economic frameworks might emerge? These questions complicate the narrative but hint at a potentially transformative phase in the industry.
Reassessing Power Dynamics
Further complicating the landscape is the observation that hardware manufacturers, such as Samsung, control the vital hardware and sensor elements that underpin digital experiences. If real-time sensor data becomes the backbone of the Data Economy, why wouldn’t these manufacturers leverage their position to reshape their dependence on software platforms?
Conclusion: A Clear Signal for the Future
The rollout of the Galaxy S26 transcends mere smartphone features; it signals a broader shift towards intent-driven computing. As technical capabilities evolve and economic incentives align, the friction between desire and fulfillment will continue to narrow. The question remains: will economic gatekeepers allow these transformative layers to dissolve? The journey ahead will surely be a captivating one to watch.