The Impact of AI on SaaS
Author: Redwood Valuation Content Team
Published: March 30, 2026
Artificial intelligence is no longer just another feature layered into software, it's beginning to reshape the economic structure of the SaaS model itself.
Over the past 18 months, venture capital has shifted meaningfully toward AI-enabled software. Industry data from 2025 shows that approximately half of all global venture funding flowed to AI-related companies, up from around one-third in 2024 [1], with much of it directed toward application-layer software. At the same time, deal counts remain well below the peaks of 2021, while mega-deals (transactions over $100 million) accounted for roughly 73% of total AI venture investment value in 2025. [2]
The result? A market defined less by exuberance and more by concentration. Fewer deals, larger checks, and higher conviction. In other words, capital is still available, but it's becoming more selective about where it goes.
Investors are gravitating toward companies that demonstrate durable advantages, whether through proprietary data, strong distribution, or measurable operational impact. Simply adding generative features to an existing product is no longer enough. The real question is whether AI materially changes how the software creates value.
From Workflow Software to Outcome Software
For most of the past decade, SaaS followed a predictable model. Software helped people work faster, companies paid per seat, and growth came through expansion and retention. AI is beginning to change that structure.
Survey data suggests that roughly three-quarters of organizations now use AI in at least one business function, reflecting a rapid increase in enterprise adoption. [3] Scaling those deployments across core workflows remains a work in progress for most, but the more important shift is what those systems are actually doing. Increasingly, they are executing the work themselves.
AI-enabled platforms can now draft agreements, reconcile datasets, detect anomalies, forecast demand, and resolve customer issues with minimal supervision. As that happens, the value of the platform starts to move away from the interface and toward the decision engine underneath it.
This shift has meaningful valuation implications. For years, SaaS performance was judged largely by recurring revenue growth, retention, and seat expansion. AI-enabled platforms operate at a deeper layer of the business, shaping how work gets done and how value is captured. When software begins to alter cost structures or accelerate revenue cycles, the way value is measured starts to change.
Capital Is Flowing, but It's Concentrated
Software financing improved in 2025 compared to the previous two years, but the distribution of capital has become increasingly uneven. A relatively small share of AI financings now captures a disproportionate percentage of venture dollars. Larger rounds tend to cluster around companies that demonstrate the following:
• Proprietary or defensible training data
• Deep specialization in a particular industry
• Evidence that the software is already embedded in enterprise operations
Horizontal SaaS platforms without differentiated data assets are beginning to face more pressure on both margins and valuation. Meanwhile, vertical AI platforms in sectors such as healthcare, legal services, financial services, and industrial operations are attracting earlier institutional capital.
Where growth once dominated investor conversations, attention is shifting toward whether a company's competitive advantage can hold up over time. Investors are increasingly asking: How embedded is the system? How proprietary is the data loop? How difficult would it be for another platform to replace the intelligence layer?
The Quiet Shift in Pricing
Another change happening beneath the surface is pricing. While traditional SaaS businesses relied on subscriptions and seat-based licensing, AI systems are starting to introduce a different model: pricing tied to output. Vendors are linking revenue to tasks completed, transactions processed, or measurable outcomes delivered. This alignment can strengthen lifetime value, but it also introduces complexity.
AI workloads are driving higher demand for cloud infrastructure, which means compute costs, especially for inference and model retraining, are now directly affecting gross margins. As a result, forecasting becomes more nuanced. Consumption-based pricing can introduce greater revenue variability, and traditional SaaS margin assumptions may not hold when computational intensity becomes central to the business.
Data Moats and Governance Risk
Historically, SaaS platforms built switching costs through workflow integration. Once a product became embedded in a company's operations, replacing it was difficult. AI adds another layer of defensibility: data feedback loops.
As models learn from proprietary customer datasets, performance improves over time. That learning cycle can strengthen retention and deepen switching costs. At the same time, it introduces new governance considerations.
Throughout 2025, several jurisdictions advanced regulatory frameworks addressing AI transparency, training-data accountability, and model risk oversight. For companies operating in regulated industries, diligence increasingly extends beyond revenue metrics to include data provenance, model governance, security architecture, and regulatory exposure.
Technical architecture is becoming inseparable from enterprise value.
Margin Dispersion Ahead
Traditional SaaS businesses benefited from strong incremental margins once infrastructure scaled, but AI is beginning to change that dynamic. AI-enabled software carries heavier compute requirements, which means margins are increasingly tied to how efficiently companies manage model performance, infrastructure, and pricing.
Scale can still support margin expansion, but the path is becoming less uniform across the sector. Companies that control inference costs and secure efficient infrastructure economics may preserve strong margins. Others may face compression if computational expenses rise faster than pricing power. As a result, investors are starting to draw a clearer line between companies where AI is simply a feature and those where it sits at the economic core of the platform.
What This Means for Founders and Investors
AI is quickly becoming standard across the software industry, and what will matter is how deeply it changes the product. While some companies will use AI to enhance existing workflows, others will use it to automate them entirely. That difference matters more than any feature roadmap because it changes how value is created and how software companies are valued.
As investors evaluate the next generation of SaaS companies, that divide will become increasingly visible. And it will likely determine which platforms command the strongest valuations in the years ahead.
[1] “6 Charts That Show the Big AI Funding Trends of 2025,” Crunchbase News, 2025. https://news.crunchbase.com/ai/big-funding-trends-charts-eoy-2025/
[2] OECD, Venture Capital Investments in Artificial Intelligence Through 2025, OECD Policy Briefs No. 50, OECD Publishing, 2026. https://www.oecd.org/en/publications/venture-capital-investments-in-artificial-intelligence-through-2025_a13752f5-en/full-report.html
[3] McKinsey & Company, The State of AI in 2025: Agents, Innovation, and Transformation, 2025. https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai