泰坦的冲突（英）2026

EMERGING TECH RESEARCHClash of the Titans Institutional Research Group Derek HernandezSenior Research Analyst,Enterprise SaaS andInfrastructure SaaS Incumbents versus challengers in the age of agentic AI pbinstitutionalresearch@pitchbook.com PitchBook is a Morningstar company providing the most comprehensive, mostaccurate, and hard-to-find data for professionals doing business in the private markets. Published on January 14, 2026 Contents Key takeaways Key takeawaysThe radical transformation in SaaSAI-embedded incumbentsAI-native challengersMarket development and investmentMarket sizing and growthOpportunities and constraintsDefensibility and the tech stackThe new moatsIncumbents and challengers meetThe investor perspectiveConclusion and outlookAppendix: Scoring key incumbents andchallengers •The enterprise SaaS sector is undergoing its most significant technological shift ina generation, driven by the maturation of artificial intelligence. This report provides aframework for private market investors to navigate this transformation, which is defined •While the enterprise adoption of AI is high (78% of organizations), meaningful businessoutcomes are not (95% of pilots are failing to accelerate revenue).1This execution gap creates a massive opportunity for a new class of vendors. •This market is bifurcating. Incumbents are embedding AI copilots into legacysuites, leveraging their vast distribution and customer trust. In contrast, AI-native •For investors, we argue that the most durable value will not come from the AI modelsthemselves, which are becoming commoditized. Instead, the new, defensible moats arebeing built on proprietary data pipelines that create a virtuous feedback loop, agentic •This report analyzes the market dynamics, investment landscape (including a $65billion TAM set to grow to $190 billion by 2030), and competitive strategies of bothincumbents and challengers. It concludes with an actionable playbook and diligence The radical transformation in SaaS The enterprise software-as-a-service (SaaS) sector is undergoing its most significanttechnological shift in a generation, even larger than the transformation fromperpetually licensed products to SaaS itself. This shift is driven by the ongoingmaturation of AI, especially the advancements by major large language models (LLMs) Thus, this is the first in a series of reports on AI within enterprise SaaS, beginning witha high-level overview of the state of play today. In our future pieces, we will dive into AIin specific sectors and subsectors, including AI in customer relationship management(CRM), AI in HR tech, and others. We invite PitchBook clients to reach out with specific The AI transformation of enterprise SaaS has sorted the segment into a criticaldichotomy: the distinction between retrofitting existing systems with AI features(AI-embedded platforms) and designing new operations from the ground up around The proliferation of generative AI has forced every enterprise software companyto formulate an AI strategy. However, not all AI is created equal. The architecturalapproach an organization takes—either building new systems with AI at their core or AI-embedded incumbents AI-embedded solutions, thus far the domain of legacy SaaS incumbents, offeradditional productivity gains within established solutions and platforms. Theseincumbents within enterprise SaaS are embedding copilots and task agentsinto existing suites, spanning CRM, enterprise resource planning (ERP), supplychain management (SCM), HR, collaboration, IT service management (ITSM),and cybersecurity. They include Salesforce’s programmatic push to Agentforce,Microsoft’s various Copilot integrations across Microsoft 365 and Dynamics 365, The AI-embedded approach keeps the user inside an incumbent system (ERP, CRM,ITSM, help desk, integrated development environments, and finance stacks) andinserts model-driven decision-making at a specific step. The difference is less However, the impact of this approach is inherently constrained by the underlyingarchitecture of the legacy systems it augments. These systems are often burdenedby technical debt, fragmented data silos, and rigid workflows, which work against thekind of transformative change clients are demanding today. Generally, AI-embeddedsoftware follows the traditional approach of building upon established systems thatusers are familiar with and enterprises are often already heavily invested in. The AI-native challengers In sharp contrast, AI-native platforms are built on a different foundation. An AI-native workflow is a business process conceived and engineered around an agent ormodel from inception. The agent plans steps, selects tools, reads from and writes to In our view, and in the general view of the market so far, it is the AI-native vendorsthat are creating a new class of agile, hyperscalable, and possibly peak-efficiencybusinesses. In an AI-native workflow, a model or agent proposes or executes steps,w