{"id":16793,"date":"2026-07-21T11:21:03","date_gmt":"2026-07-21T09:21:03","guid":{"rendered":"https:\/\/bcombinator.com\/2026\/07\/21\/the-future-of-saas-in-the-age-of-ai-agents\/"},"modified":"2026-07-21T14:48:23","modified_gmt":"2026-07-21T12:48:23","slug":"the-future-of-saas-in-the-age-of-ai-agents","status":"publish","type":"post","link":"https:\/\/bcombinator.com\/en\/2026\/07\/21\/the-future-of-saas-in-the-age-of-ai-agents\/","title":{"rendered":"The Future of SaaS in the Age of AI Agents"},"content":{"rendered":"\n
<\/div>\n\n

For more than two decades, the SaaS model has dominated enterprise software based on a simple premise: charge per user, drive adoption, and expand the number of licenses across organizations.<\/p>\n\n

However, the emergence of AI agents is beginning to challenge the economic foundations on which this industry was built.<\/p>\n\n

The question is no longer whether AI will transform SaaS, but how much value traditional software will continue to capture when agents are capable of performing the work themselves. The data already reflects the scale of this shift: according to recent figures from CB Insights, by early 2026 the software sector had lost nearly $1 trillion in market value <\/strong>following successive launches of agentic AI tools\u2014a sign that investors are reassessing the terminal value of these companies, not just their current revenues. <\/p>\n\n

<\/div>\n\n

From User-Centric Software to Autonomous Software<\/strong><\/h2>\n\n
<\/div>\n\n

The SaaS model was designed for people: one license per employee, interfaces built for human interaction, and growth tied to increasing user adoption.<\/p>\n\n

AI agents change that logic.<\/p>\n\n

They do not require individual credentials, they do not rely on complex user interfaces, and they can autonomously perform tasks such as retrieving information, generating documents, resolving incidents, or coordinating workflows across multiple systems.<\/p>\n\n

As the human user ceases to be the center of operations, the concept of the “seat” as the primary unit of monetization also begins to lose relevance.<\/p>\n\n

This shift has major implications for an industry whose valuations have historically depended on the continuous expansion of user licenses.<\/p>\n\n

<\/div>\n\n

The Challenge to the Seat-Based Model<\/strong><\/h2>\n\n
<\/div>\n\n

Much of SaaS growth has been built on a simple assumption: more employees mean more licenses and, therefore, higher revenue.<\/p>\n\n

But a single AI agent can perform work that previously required multiple employees.<\/p>\n\n

Fewer users mean fewer licenses.<\/p>\n\n

This forces software companies to rethink how value is created in a world where autonomous systems increasingly perform the work.<\/p>\n\n

As a result, many companies are experimenting with new monetization models. This is far from a niche trend: today, more than 2,000 companies offer AI agents or AI assistants, many of them built from the ground up without the constraints of legacy seat-based pricing. <\/p>\n\n

Three monetization approaches are emerging:<\/p>\n\n