Chris ChoPublished: March 25th, 2026
America's largest corporations are holding onto their core business software even as artificial intelligence threatens to upend the industry, instead using the moment to negotiate better deals with vendors and build their own AI-powered customizations on top of existing systems.
The strategy offers an early look at how enterprises plan to navigate a future where AI agents automate workflows while legacy software from companies like Salesforce, SAP and Workday continues to serve as the foundation for corporate data.
“We are not looking at this point for [software] to leave our enterprise,” said Vishal Talwar, chief digital and information officer at FedEx. But he sees the uncertainty over software's future as leverage. “We're contemplating all of these shifts and having proactive conversations with each of our partners to see how they plan to keep up.”
The software selloff
Software stocks have taken a beating in recent weeks as investors question whether traditional enterprise vendors can withstand competition from AI-powered rivals. Shares of Salesforce and Adobe have dropped sharply, with each new announcement from AI developers intensifying the selloff.
The market turmoil stems from fears that generative AI tools—which can now write code, automate workflows and build custom applications in minutes—will make expensive enterprise software obsolete. But corporate technology chiefs say that narrative misses the complexity of running large organizations.
Mike Kempe, chief information officer at audit and consulting firm Grant Thornton, said business software like enterprise resource planning systems handle regulatory requirements, multiple geographies and languages in ways that make them “very challenging to build using AI today.”
There's also the maintenance burden. Once companies build their own software, they need dedicated engineering teams to keep it updated. “I'd much rather spend our internal dollars and effort building something that is truly cutting-edge and helps us grow,” Kempe said.
The rise of vibe-coding
Still, AI coding tools have made it easier than ever to build software in-house through what's become known as “vibe-coding”—using AI assistants to rapidly prototype applications and customizations. According to Gartner, 96% of software engineering organizations now offer AI tools to their developers, and over 70% of business software developers are expected to use AI agent-based coding tools within two years.
Controlled studies show developers complete tasks up to 55% faster with AI coding assistants, while more than 75% of developers now use AI daily. Teams report productivity gains of one-third or more when AI tools are paired with strong testing and deployment practices.
Some companies are already seeing major cost savings. Thimaya Subaiya, executive vice president of operations at Cisco Systems, said the networking-equipment maker replaced a presentation software tool with its own AI agent, saving nearly $5 million annually in license costs. Cisco is now eyeing other software vendors that cost the company $50 million to $200 million each year in subscription fees.
“You need to look at all the applications that you're using and say, ‘Which one of these can become automated workflows where we don't need this application anymore?'” Subaiya said.
Ernst & Young is taking a similar approach with its $1 billion annual technology budget. The firm doesn't plan to abandon its longstanding SAP software, but it's using vibe-coding and AI agents to build customizations on top of it rather than purchasing costly upgrades directly from SAP.
“If this AI wasn't there, vibe-coding wasn't there, the agentic frameworks weren't there, it would've taken another very costly, expensive upgrade of the SAP software,” said Raj Sharma, a global managing partner for growth and innovation at EY.
Agents as the new interface
The shift points to a future where AI agents become the primary way employees interact with software, while traditional applications fade into the background as data repositories.
Cisco's Subaiya described replacing business software as replacing business processes with AI agents. “You don't need an application anymore, because an application just becomes part of the agent database,” he said.
Kempe at Grant Thornton sees core business software evolving from the heart of how businesses operate into a source of corporate data, with AI agents leading the charge as the new interface.
Seemantini Godbole, chief digital and information officer at Lowe's, said it's unlikely all employees will need a “full-fledged” customer relationship management system in the future. Instead, they'll likely need simpler, AI agent or chat-based interfaces for the software.
The home-improvement giant is already vibe-coding its own content-generation tools to manage IT spending. Lowe's distributes a circular to every store that regularly requires new image creation—a task now handled by custom AI tools rather than purchased software.
“There will be some software vendors who will be extremely innovative,” Godbole said. “But there are also others where we'll say, ‘Actually, we can do this much better with OpenAI out of the box.'”
Market transformation underway
The agentic AI market reached between $9.14 billion and $10.86 billion in 2026, according to industry estimates. By 2025, 44% of companies had deployed or assessed AI agents, with many scaling to production in 2026 for tasks like code development and finance automation.
Vista Equity Partners predicts agentic enterprise solutions will expand software markets by enabling agents to work alongside humans, creating what the firm calls “structurally superior” business models compared to traditional software-as-a-service through better workflow integration and trusted execution.
But the transformation faces organizational barriers. According to a 2026 Deloitte report, 84% of companies have not redesigned jobs or workflows for AI—a gap that positions execution-focused companies to lead while others struggle with data silos and unclear return on investment.
Worker access to AI rose 50% in 2025, and companies with at least 40% of AI projects in production are expected to double soon. Enterprises spent over $30 billion on AI pilots in 2025, dubbed “the year of the AI pilot” by industry observers.
Small vs. large companies
The divide between small and large enterprises is becoming clearer. Some small and midsize companies have successfully vibe-coded their own customer relationship management systems from scratch—something technology leaders say larger firms can't easily replicate.
Amjad Masad, founder and chief executive of coding startup Replit, said many of the company's customers are small and medium-size businesses that build their own CRMs. But larger enterprises face different constraints.
“What they want to do is create their own processes, their own workflows, their own agents, their own automations on top of it,” Masad said, referring to existing software platforms.
The complexity gap stems from regulatory requirements, global operations and legal frameworks that small companies don't encounter. FedEx's Talwar said the shipping giant doesn't view core software “as systems of differentiation,” meaning the company's strategy is to put technology resources into building tools that offer unique competitive advantages rather than recreating basic infrastructure.
Vendor response
The pressure is forcing software vendors to evolve. Analysts expect companies like Salesforce and Adobe to pivot toward agent-compatible platforms and outcome-based pricing models as AI agents commoditize user interfaces.
Enterprise software growth had already slowed after the golden era of Office 365 adoption from 2015 to 2022, with companies consolidating vendors and curbing cloud spending. AI agents could revive expansion by automating workflows across systems, but only if vendors adapt quickly.
BNY Wealth analysts noted that agentic AI represents a shift to “structural automation” that challenges traditional software moats like switching costs. The analysts expect AI budgets to increase as companies move from pilots to production.
For now, large enterprises are using the market uncertainty as leverage in vendor negotiations while building internal AI capabilities. The approach lets them prepare for an agent-driven future without abandoning the complex systems that still run their operations.
