For over twenty years, the software industry has primarily operated on the seat-based software-as-a-service (SaaS) paradigm, where customers purchase licenses for individual users to access software platforms. This established system facilitated companies like Salesforce, ServiceNow, Adobe, Snowflake, and Intuit to build extensive ecosystems around user interaction with complex software interfaces. However, this fundamental model is now facing a transformative challenge from the rise of agentic artificial intelligence technologies.
Recent developments in AI have introduced autonomous agents such as Anthropic's Claude Cowork, OpenAI's Codex, and newer solutions like Operator, which are capable of orchestrating intricate data pipelines, automating outreach, and executing multiple platform interactions without direct human supervision. This progress calls into question the sustained relevance of pricing structures based on individual software licenses, as a single AI agent can now perform tasks that historically required hundreds of human-operated licenses.
Experimentation with AI personal assistants running locally on users' machines, including projects like Clawdbot, Moltbot, and Openbot, exemplifies this shift. These opensource agents autonomously manage tasks such as file organization, email correspondence, and internet browsing. Moreover, they have been credited with developing bespoke software solutions to fulfill workflows in the absence of existing alternatives, indicating an appetite for agent-driven automation replacing manual software engagement.
Examining the impact on legacy software companies reveals three primary pressure points:
- Valuation Compression: Stock valuations for companies like Snowflake and Intuit traditionally rely on premium margins sustained by subscription licenses. However, as AI bots automate functions like accounting or generate code autonomously via application programming interfaces (APIs), the appeal and necessity of purchasing high-cost software licenses diminish, threatening these firms' revenue models.
- Reduction in Skill-Gating: Adobe, known for its Creative Cloud suite, maintains a business model where customers pay for access to sophisticated tools requiring specialist skills. Agentic AI now allows marketing teams to instruct agents to generate numerous marketing content variants and perform A/B testing automatically. This ability reduces dependence on expansive editing toolkits and the associated licensing fees.
- Erosion of the CRM Centrality: Salesforce's dominance arose from being the essential repository where organizational data resides. With AI agents capable of synthesizing data across Slack, email, PDFs, and calendars without direct user input to update CRM stages or log calls, the necessity for manual data entry and direct CRM interaction declines significantly.
These shifts suggest a potential transformation in how enterprise software functions as a broader technological infrastructure. Goldman Sachs Chief Information Officer Marco Argenti emphasized that this progression marks technology moving towards AI models acting as the new operating systems. He highlighted that agent-as-a-service could effectively replace the incumbent software-as-a-service framework. Argenti reflected on over four decades in technology, asserting that the changes observed in 2025 are unprecedented in his experience and anticipates even more profound developments in 2026.
Market responses have already manifested as declines in the share prices of key software companies. Salesforce’s stock recently experienced a significant downward movement, with a price noted at $193.91, declining over 8%. ServiceNow’s shares dropped similarly, trading near $108.53. Snowflake saw an approximately 11% decline to around $169.67, while Intuit’s pricing fell by more than 12% to just below $470. Adobe’s stock price also fell, nearly 7.5% to $282.76. These market reactions reflect investor concerns regarding the impact of AI-driven automation on traditional software business models built on subscription licensing.
In essence, the evolving AI landscape introduces both challenges and uncertainties for existing software vendors. The seat-license model’s viability is under scrutiny as AI agents promise efficiency gains, cost savings, and reduced dependence on human-operated software tools. Companies deeply invested in legacy SaaS revenue streams face risks tied to diminishing customer willingness to maintain costly licenses in favor of AI-powered alternatives. Furthermore, this paradigm shift may recalibrate valuation metrics, disrupt established customer workflows, and reshape the competitive dynamics within enterprise software markets.