AI vs. SaaS: Why Software Valuations are Under Pressure

Something I’ve been hearing a lot about recently is how software valuations appear to be weakening – no, plummeting, straight through the floor. 

Two of my clients pointed to Claude Opus 4.6 as the catalyst. This is Anthropic’s most intelligent and efficient AI model to date, and it’s joined by many others on the market. 

My clients’ rightly pointed out that if AI tools like Opus 4.6 can generate the same outputs as specialist software, for a fraction of the cost, what point would customers see in continuing to pay for multiple SaaS products? 

Over $1tn in software market value has already been lost in 2026, with AI cited as the key offender. There’s no doubt that many SaaS businesses are having their lunch eaten by AI. 

But I don’t see the underlying issue being that software demand is disappearing. It is simply that the basis of that demand is changing, and the companies who respond to that are getting to keep their lunch and eat it, too.

The model that made SaaS so valuable

For years, SaaS ran on a simple premise: sell access. Revenue came from “seats”. A company with 1,000 employees might pay £10,000 per seat annually. The cost to deliver that software was minimal, so margins were high and predictable. Meanwhile, contracts locked customers in, often encouraging them to buy more licenses than they actually needed.

That combination of recurring revenue, low delivery costs, and long-term contracts made SaaS one of the most attractive investment categories of the past decade. Double-digit EBIT multiples were common three or four years ago. Now, AI tools are starting to crack that bedrock.

AI breaks the link between access and value

Models like Opus 4.6 introduce a different way of interacting with technology. Instead of buying access to multiple software tools to perform multiple tasks, users can get the outputs instantly from one platform. Research, financial modelling, reporting, content creation, coding – these are tasks that once required separate software, but can now be generated through a single interface.

In a nutshell, pricing has moved away from seats and towards utilisation; you pay for what is produced rather than who has access. A small number of users can now achieve, through a single AI platform, what previously required entire teams spread across multiple software tools and licenses. 

This erodes one of the core drivers of SaaS valuation. If revenue is no longer tied to headcount, it becomes harder to forecast. For investors, that introduces uncertainty into a model that was built on predictability.

How SaaS valuations are being affected

The term “SaaS-pocalypse” has begun circulating in some online spheres as the scale of value erosion increases. Shares in Atlassian Corp, for example, have fallen around 50% since the start of this year. 

Likewise, Zoom’s share prices were down 20% March 2026, likely a result of the commoditisation of AI video features. As capabilities are absorbed into platforms like Microsoft Teams and Google Workspace, standalone video tools like Zoom look increasingly expendable. 

Investors are also questioning the long-term defensibility of legacy risk and compliance SaaS platforms, whose value lies in monitoring, flagging and reporting – exactly what AI replaces best. And why would you pay for bespoke accounting, sales analytics, logistics, or project management software when you could just ask ChatGPT, Claude, or Gemini to do it?

Valuations have adjusted in response. Pandemic-era SaaS multiples of around 9.8x revenue have compressed into the 4.3-5.3x range, reflecting a dire reassessment of how durable these revenue streams will be.

But there are companies going in the other direction

While many SaaS businesses face pressure, others are seeing their valuations strengthen.

• Microsoft has embedded Copilot across its core products, turning its ecosystem into a central layer for AI-enabled work.
• Salesforce has integrated Einstein AI into its CRM platform, extending its role into forecasting and automation.
• BlackRock’s Aladdin has applied AI to investment decision-making, reinforcing its position within client workflows.

In each case, AI is being used to deepen the value of the product, and the market rewards that. AI-integrated SaaS companies are achieving 15-24% higher valuation multiples, and 80% of PE and strategic buyers are actively paying more for AI-native SaaS.

A divide is opening up. On one side sits companies whose offerings can be replicated or substituted by general-purpose AI, which weakens their pricing power. On the other side are businesses that integrate AI into their core architecture, which strengthens their position within client workflows. These companies attract capital at higher multiples because their narrative aligns with future demand. 

A decision point for SaaS leaders

AI is altering the criteria investors use to judge SaaS businesses. Predictability, once anchored in seat-based pricing, carries less weight. Relevance, measured through integration and impact on outcomes, carries more. 

SaaS leaders now face a clear choice about how they position their business. Those that continue to rely on access-based pricing models will find it harder to defend their value. But those that integrate AI into the core of their product, and use it to improve their value proposition, will strengthen their relevance in tomorrow’s world and justify higher valuation multiples as a result.

The opportunity is still there for businesses to seize. The AI SaaS market is projected to grow at ~38% CAGR, taking it from $71bn today towards $775bn over time. Based on that, I’d say demand isn’t disappearing, but it is concentrating around the companies that know how to apply AI in a way that actually drives outcomes.