AI Ends the Era of Traditional Software Licensing: A Radical Shift in the Tech Industry

The software sector is undergoing a radical transformation thanks to the rise of Artificial Intelligence, with major companies like OpenAI and Anthropic moving to change their pricing models. This shift is expected to lead to the demise of seat-based software licenses, which have been the prevailing standard for decades.

The Shift Towards Results- and Consumption-Based Pricing Models

A McKinsey analysis predicts that the "post-SaaS" era has already begun. With increasing interactions between AI agents, software and services will be delivered in new ways, leading to a decline in the per-seat pricing model.

Instead, new models will be based on achieved results or consumption volume. According to the McKinsey team, led by Jeremy Schneider, this fundamental shift redefines what software is, who builds and uses it, and how companies are organized and operated.

Implications of the Shift for Vendors and Buyers

For software buyers and users, this necessitates a restructuring of their relationships, even with their preferred vendors.

Schneider and his colleagues expect these developments to lead to an acceleration in vendor changes and an increase in customer churn rates, in addition to a reorganization of user segments and value pools, and an increase in in-house or citizen-developer-led software development.

AI Drives Growth and Reduces Costs

A growing percentage of software vendors are betting on an AI-driven future, with 40% expecting AI to unlock revenue growth exceeding 20% above their current trajectory, while 11% anticipate revenue gains exceeding 50%.

They also expect significant reductions in operating costs.

Faster Evolution than Cloud Computing Adoption

There is a general consensus in the industry that the software delivery model is undergoing a significant transformation, driven by AI and AI agents.

George Brown, a partner at Partner Economics, predicted that "the transition to an agent-based architecture will evolve much faster than the evolution of client/server-based applications to cloud computing" (ZDNET).

He added that companies cannot ignore cost reductions of more than 20%, as this would make them uncompetitive.

Additional Link: For more information on Quantum Computing, you can visit this link.

New Pricing Models on the Horizon

The most obvious change for users will be how they acquire and pay for software – a consumption-based model for software licenses, versus the traditional per-seat model.

Nearly two-thirds of vendors in McKinsey's survey (63%) believe that AI will "fundamentally change their business model in the next three to five years."

Consumption-based pricing, based on usage, tokens, outputs, or results, will replace traditional per-seat or per-user pricing.

Selling Results, Not Licenses

Ali Gowher, Head of HR at Software Finder, says we are "on the verge of an industry where software companies need to sell results, not licenses."

Customers must now demand that agreements are based on value versus performance.

Charging the same price to someone who logs in once a week and someone who uses the service 12 hours a day makes no sense.

Usage rates give companies the opportunity to scale properly and get better deals.

The trick here is to understand your usage patterns before investing.

Agents Will Drastically Reduce Seat Count

Brown expects that software transactions between AI agents "will reduce seat counts by up to 70%."

However, he does not see selling tokens, compute time, or agent development time as viable monetization methods, because these methods are "difficult for the average businessperson to understand."

This will shift to results-based payments or specific lease costs for agents for specific, time-bound tasks.

Changing How Software is Used and Valued

This will lead to a fundamental shift in how software solutions are used, evaluated, and integrated into operations, says Natasha Crissafulli, Executive Director of Tax and Accounting Products at Wolters Kluwer.

AI agents will increasingly act on behalf of users, automating tasks, making decisions, and collaborating across systems.

This reduces the need for direct human interaction with software interfaces.

Importance of Human Oversight and Ethics

Most importantly: Crissafulli added, "As AI agents take on more responsibility, proper human oversight, transparency, ethical use, and data governance will become critical."

Customers will need to ensure that systems operate reliably and responsibly.

Risks for Customers and the Need for Verification

Ofir Bloch, VP of Strategic Positioning at WalkMe, warns that the risk to software customers is that vendors might "repackage bundle pricing as results bundles, betting you won't notice."

You could end up with closed bundles and measurement systems that you cannot audit.

Important tip: Before signing anything consumption-based, ask to see performance data and test individual capabilities.

Do not commit until you verify the vendor's ability to measure what they are charging you for.

These systems still break down when conditions change.

It's workflow automation, not true agent autonomy.

Not an Entirely New Concept

McKinsey analysis: According to McKinsey authors, the idea of replacing per-seat licenses with results-based models is not necessarily a new concept.

Between 2015 and 2024, the number of consumption-based software companies doubled, and leaders like Salesforce, Zendesk, Intercom, and LexisNexis have already monetized their AI capabilities through these models, often achieving significantly higher revenue per customer compared to traditional units.

Challenges of Transition and Future Development

It may take some time for the rest of the industry to evolve and adapt.

Expert opinion: Rob Fuller, Chief Solutions Officer at Blend360, said that "the set of technologies needed to fully realize this vision is not yet ready."

Any model involving internal agents interacting with external agents is very immature, and at this stage, humans using agents is more common and practical, with some narrow and intentional exceptions.