AI Won’t Kill Software. Here’s How To Hunt For Value

Earlier this month, nearly $300 billion in software market value was lost in a single trading day after Anthropic rolled out Claude Cowork and its open-source plugins that let AI agents execute complex tasks autonomously from raw inputs. Live demos of end-to-end legal research and drafted filings rattled investors, crushing software, data services, and IT outsourcing stocks. So is this the beginning of the end of software?

No. The narrative that AI will render traditional software obsolete is an oversimplification. Granted, AI is rapidly changing how code is created and tasks are executed, but it is reshaping the software landscape rather than replacing it entirely. When markets overgeneralize, capital often moves indiscriminately. That creates mispricings, and there may be money to be made. The key for investors is to distinguish between software companies with durable competitive advantages and those at risk of significant disruption.

Why Critical Software Survives

When AI agents began performing complex tasks like legal research and drafting documents, investors quickly feared that traditional software might become obsolete.

But that reaction overlooks a key distinction: carrying out a task is not the same as running a business system. The software most likely to endure is what companies call a “system of record,” which is the official source for critical data such as financial transactions, customer histories, payroll, or compliance records. These platforms store years of structured information and are deeply embedded in daily operations. Many enterprise systems function as the foundation of a company’s operations. Firms may upgrade tools and automate workflows, but they do not rip out the foundation of a skyscraper while the building is still in use.

Software also anchors complex workflows, coordinating multi-step processes across departments from approval chains to billing and reporting. Replacing them is costly, risky, and operationally disruptive, creating strong switching costs.

In addition, enterprise systems manage permissions, approvals, audit trails, and regulatory requirements that generic AI models cannot handle on their own. AI can automate tasks within these systems, but it still depends on them for secure data, workflow control, and governance. In many cases, AI enhances core software rather than replaces it. Related: Anthropic’s Big Software Reset: Winners & Losers

Pockets of Value: Where to Invest

Investors should focus on sectors where AI acts as an accelerator or where specialized domain knowledge provides a defensive moat.

• Vertical SaaS: This term refers to “Software as a Service” companies tailored for specific industries. These companies embed deep, industry-specific expertise and regulatory compliance into their products. For example, software designed for managing clinical trials in life sciences or handling compliance in banking incorporates complex rules that general AI models cannot reliably replicate without significant legal risk. Some companies to look at include Veeva Systems (down 20% year-to-date), which is built specifically for pharma and life sciences focused on clinical, regulatory, and quality workflows, and Guidewire (-32% year-to-date), a core system for insurers.
• Cybersecurity: AI is a double-edged sword; it helps defenders but also enables attackers to launch more sophisticated threats faster. This dynamic creates a permanent tailwind for cybersecurity software. Companies that provide essential defense infrastructure to protect networks, data, and applications are becoming more critical as the threat landscape expands. Some companies that could benefit include Palo Alto Networks (-7% year-to-date), CrowdStrike (-5% year-to-date) and Zscaler (-20% year-to-date)
• Data Infrastructure: AI models are only as good as the data they are trained on. They require vast amounts of clean, organized, and accessible data to function. Companies that provide the fundamental platforms for storing, processing, and managing corporate data are essential for the AI build-out. Companies like Snowflake (-16% year-to-date) and Oracle (-18%) fit this bill.
• Horizontal Workflow Platforms: Companies that run core business workflows across industries can also benefit from AI rather than be hurt by it. AI tools still require structured data, access controls, and built-in processes to work well. With deep integrations and high switching costs, platforms such as Salesforce (-25% year-to-date) and ServiceNow (-28%) are positioning themselves as control centers for business. Switching costs are also substantial. Sales teams run their daily operations inside these systems. Data histories stretch back years. That being said, the key question is whether AI increases revenue per customer faster than it reduces software seats.

Pockets of Risk: What to Avoid

On the other hand, business models that rely on simple tasks or generic information are highly vulnerable to AI disruption.

• Wrapper Applications: These are products that essentially offer a user interface built on top of another company’s foundational AI model (such as OpenAI’s GPT). They often lack proprietary technology and are easily rendered obsolete as the underlying model providers improve their own direct offerings.
• Seat-Based Pricing in Efficiency Sectors: Many software companies charge based on the number of human users, or “seats.” In functions where AI can significantly boost individual productivity – such as basic code generation, entry-level copywriting, or tier-one customer support – companies may need fewer humans to do the same amount of work. This leads to a reduction in the number of software licenses purchased, directly impacting revenue. Some names to be cautious of include Freshworks and HubSpot.
• Commoditized Knowledge Platforms: EdTech or information services that provide basic answers or academic assistance are being displaced by free, powerful AI models that can provide this information instantly. Stocks at risk include Chegg, Coursera, and Udemy.

Overall, it’s safe to assume that the software sector is transitioning, not ending. Value is shifting from generic tools that aid human productivity toward platforms that govern critical data, manage complex regulated workflows, and secure the digital enterprise. Investors should prioritize companies that own proprietary data and solve intricate, high-stakes business problems that lie beyond the capabilities of general-purpose AI.

However, if you are not comfortable with stock picking, you should consider a multi-asset portfolio approach.

Multi-Asset Portfolios Are The Right Way To Protect & Grow Wealth

Inflation and market volatility can erode purchasing power over time. A dedicated wealth strategy helps you diversify beyond equities to protect and grow capital.

What if you took advantage of the current commodity super-cycle? Is a portfolio of 10% commodities, 10% gold, and 2% crypto, in addition to equities, likely to return more in the next 1-3 years? We’ve crunched the numbers. Our wealth management partner manages exactly these kinds of complex multi-asset strategies, blending real assets with high-performance equity sleeves like the Trefis High Quality Portfolio, which has returned > 105% since inception.