A seismic shift is underway in the software industry, as artificial intelligence disrupts the long-dominant SaaS (software-as-a-service) business model. Investors and founders alike are witnessing what some are calling the "SaaSpocalypse"—a wave of anxiety and market correction driven by AI’s ability to replicate, and in some cases surpass, traditional SaaS functionality. At the heart of the disruption is the rise of AI agents like Claude Code and OpenAI’s Codex, which can write, deploy, and even maintain software with minimal human input. This has fundamentally altered the "build versus buy" decision for companies. Instead of purchasing expensive SaaS licenses, businesses can now use AI to build custom tools in-house—often faster, cheaper, and more tailored to their needs. Lex Zhao of One Way Ventures noted that this shift is eroding the foundation of SaaS: the per-seat pricing model, which relies on the number of users accessing a platform. When a single AI agent can perform the work of an entire team—pulling data, generating reports, managing workflows—the need for multiple user licenses diminishes. As Abdul Abdirahman of F-Prime explained, this undermines the predictability and scalability that made SaaS so attractive. With AI capable of replicating not just core SaaS functions but also add-on features, vendors lose their ability to upsell and grow revenue from existing customers. The stakes are high. In late 2024, Klarna made headlines by replacing Salesforce’s CRM with its own AI-driven system. This wasn’t an isolated case—many companies now see building their own solutions as a viable alternative. The result? A sharp decline in SaaS stock prices. In early February, a wave of investor sell-offs wiped nearly $1 trillion from software and services markets, followed by another billion later that month. Analysts have dubbed the phenomenon FOBO investing—fear of becoming obsolete. Yet not all investors see doom. Aaron Holiday of 645 Ventures views the turmoil as a natural evolution, likening it to a snake shedding its skin. The SaaS model isn’t dying, he argues, but transforming. The real winners will be those who adapt quickly, integrating AI into their offerings in meaningful ways rather than just slapping on AI features as a marketing gimmick. The market is already reacting. When Anthropic launched AI tools like Claude Code for cybersecurity and Claude Cowork AI for legal work, stocks in traditional software companies dropped. This reflects a broader concern: if AI-native startups can deliver better, cheaper, and faster solutions, why pay for legacy SaaS platforms? AI-native companies are experimenting with new pricing models. Some charge based on usage—measured in tokens—while others are testing outcome-based pricing, where fees depend on performance. Sierra, a customer service AI startup led by former Salesforce CEO Bret Taylor, has already hit $100 million in annual recurring revenue in under two years using this model. Still, uncertainty remains. The market lacks clear evidence that new business models will deliver sustainable value. Meanwhile, SaaS IPOs are on pause. Crunchbase data shows no venture-backed SaaS companies are planning to go public anytime soon, as late-stage firms like Canva and Rippling face pressure from volatile markets, high expectations, and the fear of public scrutiny. Many experts believe companies will stay private longer, waiting for clearer signals from the first AI-native IPOs—possibly from OpenAI or Anthropic later this year. Ultimately, the future likely lies in a hybrid model. While AI will replace many SaaS functions, enterprises will still need durable, compliant, auditable software for core operations. The most resilient companies will combine AI innovation with the fundamentals of retention, margins, and defensibility—proving that even in the age of AI, solid execution still matters.