The AI revolution is shaking up the software industry. Which companies will get stronger?

Key Insights

• The artificial intelligence (AI) revolution widens the range of outcomes for software companies. Some will emerge stronger; others may face disruption.
• Autonomous driving moving from hype to reality would open a large market with attractive economics, making it a compelling AI use case.
• Mission‑critical software companies may be better positioned to benefit from AI‑driven cost savings and revenue‑boosting AI features.

The software industry is home to some of the more popular business models that have emerged over the past two decades, as the shift to cloud delivery and subscriptions has led to strong growth, high margins on incremental sales, and increased recurring revenue.

However, the prospect of artificial intelligence (AI)‑driven transformation is in the air.

Navigating these changes requires a thoughtful approach. Differentiating between the software companies that are best positioned to reap the rewards of AI and those that could find their business models disrupted will be critical.

AI creates opportunities for select software incumbents

The rise of AI could give software vendors opportunities to expand their margins while also growing their revenue.

• Cost lever: Copilots and other tools are making software developers significantly more productive, enabling companies to reduce their costs while potentially delivering product enhancements and upgrades more quickly.
• Revenue driver: Launching AI‑related features could create opportunities for software companies to increase their sales growth.

The consensus view holds that incumbent software companies should enjoy a right to win because of their distribution advantages and the potential edge that comes from being embedded in customers’ data and processes.

There is a kernel of truth to this thesis. The investment needed to structure legacy data properly and concerns about security and privacy are likely to govern how quickly and broadly enterprises adopt AI. These gating factors favor incumbents, especially if they are innovating rapidly on the AI front.

Nevertheless, the AI revolution widens the range of outcomes to the upside and the downside for established software vendors. Some companies will emerge stronger; others could prove more prone to AI‑related disruption or execution risk.

AI innovation could intensify competition for some software vendors

Software companies are likely to face intensifying competition, potentially on multiple fronts.

• AI natives: New rivals are emerging from the universe of start-ups that aren’t encumbered by technological debt, both in terms of legacy systems and the need to protect an existing business.
• Big tech: The mega‑caps building out AI infrastructure and the companies pushing the envelope on highly capable large language models could also become formidable competitors.
• Self‑serve: The broad capabilities of advanced AI models also make it easier for larger, well‑resourced enterprises to develop their own solutions, particularly for less complex use cases.

We expect potential disruptors to gravitate toward larger profit pools and addressable markets.

So far, competitive pressures have been most pronounced for companies that specialize in software used by customer service representatives—an area where AI has made meaningful inroads.

“…the AI revolution widens the range of outcomes to the upside and the downside for established software vendors.”

The development of agentic AI that can complete assigned tasks independently may also influence the level of disintermediation risk to established software companies.

Will a network of specialized task‑ or industry‑specific AI agents capture more of the value? Or will more accrue to a super AI agent that functions like an operating system, pulling data from throughout the enterprise and helping to route tasks to the most appropriate model?

Our view favors industry‑specific AI agents at this point. Most jobs, and the people who perform them, are highly specialized. Human resources representatives, for example, do not need broad data access outside of the software platforms that they use for their day‑to‑day tasks.

AI could lead to shifts in software monetization

The extent to which AI‑driven efficiency gains result in head count reductions at customers could also create uncertainty for some software companies.

Selling subscriptions based on the number of users emerged as the preeminent model in the cloud era. This approach could require a rethink in the age of AI, leading to new risks and opportunities.

• Revenue headwind: Fewer users per enterprise customer could be a headwind for revenue growth at some software companies.
• Strategic shifts: Selling AI features or offering more expensive firmwide licenses could be options, as could charging customers based on usage levels.

How these pricing models evolve and their implications for companies’ earnings and free cash flow are open questions, especially in end markets where AI‑driven productivity gains could allow customers to do more with significantly fewer employees. Tools commonly used by software developers, for example, are one area where this concern has started to emerge as AI‑driven efficiencies have taken hold.

On the other hand, shifting pricing models could unlock meaningful value for some software companies. We are already hearing of examples in enterprise software where the customer’s labor cost savings create headroom for a consumption‑based sales model to increase the vendor’s revenue beyond that of a traditional seat license.

Autonomous driving is revving up

In terms of disruptive AI growth stories, the potential for autonomous driving to move from hype to reality over the next few years would open a large addressable market with compelling economics.

The opportunity in long‑haul trucking strikes us as one of the strongest AI use cases to emerge so far. A shortage of drivers—a challenge that is likely to worsen over time—has led to cost inflation. In our view, the potential cost savings and safety improvements that autonomous‑driving software could offer versus human drivers make a compelling value proposition to customers.

“The opportunity in long-haul trucking strikes us as one of the strongest AI use cases….”

Competitive intensity in the highly fragmented trucking industry also appears favorable relative to the robotaxi model. Not only are several companies pursuing robotaxis, but the dominant rideshare applications could also put up more of a fight to defend their market share.

AI opportunity: Highly sticky software companies

We also favor software businesses that could be harder for insurgents to disrupt, should be able to enjoy AI‑driven cost savings, and appear well positioned to boost revenue by selling solutions that automate workflows for customers.

Software vendors that meet these criteria tend to exhibit certain characteristics that could offer an appealing combination of resilience and growth as the AI revolution accelerates.

• Mission‑critical moat: Software that’s at the heart of customers’ workflows, especially if it’s underpinned by significant industry knowledge and proprietary data, could be challenging for disruptors to replicate.
• Know your niche: Companies that provide software to smaller industries aren’t as likely to emerge as top targets for disruptive AI technologies. In these narrow markets, customers may also be more inclined to rely on incumbents for productivity‑enhancing AI functions, particularly if the user base skews toward small to mid-size businesses.
• Systems of action: AI’s demonstrated strength in data retrieval, summary, and analysis suggests that database‑centric products could be more at risk of disruption than systems of action.

Vendors of established software solutions used heavily in product and industrial design, for example, check all three of these boxes to varying degrees. These companies have the potential to use AI to drive meaningful efficiency gains for customers by automating some aspects of the design process and through agents that notify engineers of maintenance needs or changes in requirements.

Of course, the potential for AI‑related upside isn’t the be‑all and end‑all. Ideally, these well‑positioned software companies should also trade at reasonable valuations and offer exposure to idiosyncratic upside drivers, such as strategic shifts that could unlock value or potential improvements in how the management team allocates capital.

AI investing favors real intelligence

As AI‑driven innovation and disruption accelerate, a forward‑looking investment approach that’s rooted in a deep understanding of individual companies and industries may have a leg up on portfolios that are based solely on market capitalization.

When it comes to software investing, identifying the highly sticky application providers that can benefit from AI on the top and bottom lines will be important. The same goes for avoiding software companies at risk of disruption.

Article by T.RowePrice