ASX and global software stocks have pulled back sharply as investors reassess whether artificial intelligence is a tool that enhances software, or a threat that replaces it entirely.
Over the past few months, software share prices have sold off across the globe as fast-moving AI tools reignited concerns about the long-term durability of traditional software models. The reaction has been broad, and at times blunt. High-quality businesses and weaker ones have often been treated the same.
That creates confusion for investors. Is this a genuine structural threat, or a valuation reset driven by uncertainty and fear?
To answer that, it helps to separate AI narratives from business realities.
Why software stocks sunk all together
The trigger was not earnings. It was confidence.
New AI tools like Claude Cowork and OpenClaw have demonstrated an ability to automate tasks that once required specialised software and skilled labour. That led investors to ask an uncomfortable question. If AI agents can handle end-to-end workflows, why pay recurring licence fees for traditional software?
Markets dislike uncertainty more than bad news – particularly when valuations are lofty. When the long-term shape of an industry becomes harder to forecast, valuations compress. That is exactly what we have seen across software.
Importantly, this sell-off has happened before any widespread evidence of revenue loss, customer churn, or broken products. What has changed is sentiment, not necessarily fundamentals.
How to assess AI risk properly
Rather than asking whether AI is disruptive in general, investors are better served asking two specific questions on a company-by-company basis.
First, is AI a real threat to the company’s product, model, and customer use case?
Does the software sit at the core of operations, or does it simply automate tasks that could be replaced? Is it a system of record, or a layer that could be bypassed?
Second, has the valuation pullback gone too far relative to business quality?
Has uncertainty pushed prices well below what long-term fundamentals justify?
This framework helps filter fear from fact.
Technology One and deeply embedded systems
Technology One Ltd (ASX: TNE) provides enterprise software to governments, councils, and universities. These organisations rely on its systems for finance, human resources, asset management, and student administration.
AI may improve reporting, analytics, or workflows inside these platforms. It does not remove the need for the underlying system. Switching costs are high, contracts are sticky, and the software is deeply embedded in daily operations.
The key risk to watch is not AI headlines. It is customer churn. As long as customers remain, the business model remains intact.
Pro Medicus and mission-critical software
Pro Medicus Ltd (ASX: PME) operates in medical imaging, where reliability, speed, and accuracy are non-negotiable. Its Visage platform underpins radiology workflows at major hospitals globally.
AI will increasingly assist image interpretation. That likely increases the value of fast, high-quality imaging platforms rather than replacing them. Hospitals do not experiment lightly with core clinical infrastructure.
The recent pullback reflects valuation sensitivity and broader sector fear, not a collapse in demand or competitive position.
Gentrack and regulatory complexity
Gentrack Group Ltd (ASX: GTK) provides billing and operational software to energy and water utilities. These systems operate within complex regulatory frameworks and manage revenue collection at scale.
AI may optimise components of these processes. Replacing them entirely would require rebuilding regulatory logic, integrations, and trust. That is not trivial.
For businesses like Gentrack, AI is more likely to be layered into existing platforms than to sweep them away.
So is this an opportunity?
That depends on discipline.
I think at this stage AI will disrupt weak software. It will pressure margins in some areas. It will compress moats built only on technical complexity. Those risks are real.
But history shows that when uncertainty peaks, markets often overshoot. High-quality software businesses with embedded positions, valuable data, and strong customer relationships tend to adapt rather than disappear. This is especially true when the “switching costs” not only include vast initial sums of money, but also the cost of risk in changing from exisiting software to a new AI solution outweighs the benefits.
Periods like this are uncomfortable. They are also where long-term thinking matters most.
