Nvidia CEO and Analysts Reject Theory That AI Will Kill SaaS
Industry leaders argue that while AI lowers the cost of coding, the fundamental value of SaaS lies in business process management and stability.
The narrative that generative AI will render Software-as-a-Service (SaaS) obsolete is facing a sharp rebuke from industry leaders. Nvidia CEO Jensen Huang and prominent analysts argue that the ability to generate code does not eliminate the need for the structured platforms that manage enterprise operations.
Jensen Huang described the belief that AI will kill the SaaS model as "the most illogical thing in the world." This pushback responds to a market trend where SaaS stocks have tumbled amid fears that Large Language Models (LLMs) can prompt full-stack applications into existence, potentially bypassing traditional vendors. However, proponents of the SaaS model argue that software is fundamentally about business processes and organizational stability rather than the mere act of writing code.
The Shift to Extensible Software
Rather than replacing platforms, AI is enabling a transition toward "extensible software." In this model, customers use AI to build custom additions on top of stable SaaS cores, allowing for hyper-personalization without compromising the underlying infrastructure.
Shopify has already demonstrated this shift in practice. Following the release of its Sidekick tool, merchants created nearly 4,000 custom apps within the first three weeks. These users described the tools they needed, and the AI coded the applications using Shopify's existing API. This suggests that AI acts as a bridge between a user's business need and the platform's capabilities, rather than a replacement for the platform itself.
The Value of Business Logic
Analysts suggest that the primary barrier to software adoption is not the cost of development, but the difficulty of conceptualization. Benedict Evans noted that "the hard part is knowing that you need a tool for this in the first place and then knowing what the tool should do."
Because AI cannot independently identify organizational gaps or manage the human element of change management, the core SaaS provider remains essential. The platform provides the necessary guardrails, security, and integration that a standalone, AI-generated script cannot offer. Consequently, AI may make individual features less valuable to sell as standalone products while simultaneously making the host platform more valuable to own.
Industry Implications
This evolution increases the "stickiness" of major platform providers. When customers build their own custom extensions via AI, they become more deeply embedded in the vendor's ecosystem. The vendor maintains the core infrastructure while the customer gains the niche functionality they require without waiting for a formal product update.
However, this shift creates internal tension for software companies. While a CEO may prioritize overall platform growth and ecosystem lock-in, individual product divisions may find it harder to sell premium, narrow add-on packages if customers can generate "good enough" versions of those tools using LLMs.
What to Watch
As the industry moves forward, the focus will shift from whether AI replaces SaaS to how platforms integrate AI-driven customization. The key metric for success will be whether vendors can successfully monetize the infrastructure that hosts these AI-generated extensions, or if the democratization of coding eventually erodes the pricing power of the core platforms.