The brutal sell-off that has rattled software stocks this year may have overlooked a crucial reality: the industry’s old guard is not being outpaced by artificial intelligence, but is rapidly repositioning to profit from it.
That was the core message from OpenAI Chief Operating Officer Brad Lightcap, who has offered a forceful rebuttal to the market narrative that AI agents and custom-built tools will hollow out the traditional software business.
Speaking on the latest episode of the Uncapped podcast, Lightcap argued that established software makers, far from being caught flat-footed, are moving with the urgency of startups while leveraging advantages that younger AI-native firms can only aspire to build.
“All of these companies are as motivated and moving as quickly as any startup,” he said, adding that their long-standing customer relationships remain a powerful moat in an increasingly crowded AI market.
His assessment lands at a pivotal moment for the technology sector. Since early February, investors have aggressively repriced software stocks amid fears that generative AI and autonomous agents could displace traditional enterprise tools. The correction, widely dubbed the “software apocalypse,” has knocked major names such as Salesforce, Microsoft and Snowflake sharply lower, with several names down between roughly a quarter and a third this year.
The anxiety has been driven by a broader market thesis: if companies can increasingly use AI to build bespoke internal tools, demand for expensive software subscriptions may weaken.
Yet Lightcap’s intervention suggests the market may be underestimating how deeply incumbents are already embedding AI into their products and operations.
From OpenAI’s vantage point, working closely with large enterprise vendors, he said, these companies are not simply adding AI features as cosmetic upgrades. Rather, they are rethinking the entire customer journey, from onboarding and workflow automation to expansion into adjacent business lines.
The current debate on Wall Street is no longer whether AI will reshape software, but whether it will destroy legacy vendors or strengthen them. Lightcap clearly belongs to the latter camp.
His view is that incumbents enjoy structural advantages that remain difficult for startups to replicate quickly: entrenched enterprise contracts, access to proprietary customer data, global sales infrastructure, and trusted relationships with chief information officers and procurement teams.
In effect, AI may become less a disruptive force against these companies and more a catalyst for product reinvention. That perspective has also found support across the wider technology ecosystem.
Dan Rogers, the chief executive of Asana, argued that AI agents actually increase the need for workflow software rather than diminish it.
“With AI and AI agents, the coordination problem doesn’t go away. It actually expands exponentially,” he told BI, noting that organizations will need systems capable of managing collaboration not only between employees but also between thousands of machine agents.
That argument cuts to the heart of enterprise software’s enduring relevance. Even as AI takes over repetitive clerical functions, businesses still require the architecture to govern permissions, approvals, compliance trails, resource allocation, and task visibility. Those layers have traditionally been the preserve of established software vendors.
The same logic has been echoed by Jensen Huang, whose NVIDIA sits at the center of the AI boom. Huang dismissed the idea that software tools are in structural decline, arguing instead that AI systems will rely on existing platforms rather than replace them wholesale.
“There’s this notion that the tool industry is in decline and will be replaced by AI,” Huang said, explaining how AI will use the tools software offers and not reinvent its own.
He added, “It is the most illogical thing in the world, and time will prove itself.”
There is also a more practical cost argument underpinning the bullish case.
Anish Acharya of Andreessen Horowitz recently argued that rebuilding core business systems such as payroll, enterprise resource planning, and customer relationship management software with AI would yield only limited savings, estimated at about 10%.
“You have this innovation bazooka with these models. Why would you point it at rebuilding payroll or ERP or CRM,” Acharya said.
That suggests the economics of replacing deeply integrated systems may be less compelling than markets initially assumed.
However, the bigger issue for investors may be whether the sell-off has gone too far. Lightcap’s suggestion that being bullish on AI should also imply being bullish on legacy software amounts to a contrarian call against one of the year’s most crowded market trades.
If AI ultimately acts as an accelerant for incumbent platforms rather than a wrecking ball, the sharp markdown in software valuations could, in hindsight, look less like a rational repricing and more like an overreaction driven by short-term fear.
What is becoming increasingly clear is that the next phase of the AI race may not be defined solely by startups and frontier labs. The established software giants, armed with capital, customers, and distribution, appear determined to remain central players in the new technology cycle.
