Databricks CEO Ali Ghodsi says Software-as-a-Service (SaaS) isn’t disappearing—but its dominance may soon fade. Speaking at a recent industry event in early 2026, Ghodsi argued that rapid advances in artificial intelligence will fundamentally reshape how software is built, sold, and used. The shift, he suggested, will move enterprises away from traditional app subscriptions toward AI-driven, data-native systems. For businesses and software vendors alike, the implications could be profound.

Background

For over two decades, SaaS has been the defining model of enterprise software. Companies replaced on-premise tools with cloud-hosted subscriptions for CRM, HR, finance, and collaboration.

But the rise of generative AI and large language models has introduced a new paradigm: software that doesn’t just store and process data—it reasons over it. Vendors are now embedding AI copilots, agents, and automation layers directly into workflows.

Databricks, long focused on data lakes and analytics, has been positioning itself at the center of this transformation through its “data intelligence” platform strategy.

Key Developments

Ghodsi’s central claim: SaaS applications, as standalone systems of record, risk becoming commoditized.

Instead of logging into separate tools, he envisions AI systems that:

• Pull data from multiple enterprise sources in real time
• Generate insights automatically
• Execute business processes through AI agents
• Replace rigid dashboards with conversational interfaces

In this model, value shifts from the application layer to the data and intelligence layer.

Ghodsi emphasized that enterprises increasingly want outcomes—not software seats. AI can deliver those outcomes faster by automating analysis, reporting, and decision support.

He noted that many SaaS vendors are already racing to retrofit AI features, but warned that incremental add-ons may not be enough to compete with AI-native platforms.

Technical Explanation

Traditional SaaS works like renting a finished apartment—you use predefined rooms and layouts.

AI-native software is more like living in a smart, modular space that reconfigures itself based on your needs.

Instead of:

• Static forms
• Fixed dashboards
• Manual queries

AI systems can:

• Interpret natural language
• Generate custom reports instantly
• Automate workflows end-to-end

This is powered by large language models, vector databases, and unified data platforms that allow AI to reason across structured and unstructured enterprise data.

Implications

For enterprises:

Companies may reduce reliance on dozens of separate SaaS tools, consolidating workflows into AI platforms.

For SaaS vendors:

Pricing models based on per-seat subscriptions could erode, replaced by usage- or outcome-based pricing.

For workers:

Routine software navigation may decline as AI agents handle tasks like data entry, reporting, and analysis.

For the tech economy:

Spending could shift from application licenses to AI infrastructure and data platforms.

Challenges

Despite the bold vision, several hurdles remain:

• Data readiness: Many enterprises still struggle with siloed, low-quality data.
• Security & governance: AI agents accessing sensitive systems raise compliance risks.
• Reliability: Hallucinations and model errors limit full automation.
• Cost: Running large AI models at scale remains expensive.

Critics also argue that SaaS will evolve rather than fade—embedding AI while retaining app-centric workflows.

Future Outlook

The near future likely brings a hybrid phase:

• SaaS apps infused with copilots and agents
• AI platforms integrating multiple SaaS data streams
• New startups building AI-native enterprise tools from scratch

Over time, if AI agents become reliable operators of business processes, the interface layer—the hallmark of SaaS—could diminish in importance.

Databricks is betting that data platforms, not applications, will be the new enterprise control plane.

Conclusion

SaaS isn’t dying—but its center of gravity is shifting. As AI transforms software from passive tools into active collaborators, the value stack is moving toward data and intelligence. Whether SaaS adapts or gets abstracted away will define the next era of enterprise computing.