The chief executive of one of the world's largest enterprise software companies says human software development may soon cease entirely inside his organisation
It may seem counterintuitive at first, but in an ironic twist it may be certain tech workers who will see there roles disappear in the face of new tech. Christian Klein, chief executive of SAP, has told the Australian Financial Review that the rise of vibe coding - AI tools that allow non-technical users to generate software from plain-language instructions - means the company could have no human software developers within three to four years.
"Software development is the function most impacted by AI," Klein said, "and there is a chance that in three to four years there is actually no one developing software inside SAP any more."
SAP employs more than 110,000 people globally and is Europe's largest software company, with a market value of around $195 billion. Its shares have fallen 34% so far in 2026, part of a broader sell-off in software-as-a-service stocks as investors question the long-term viability of companies whose core products can increasingly be replicated by AI platforms.
Klein was quick to add that a smaller developer headcount need not mean fewer employees overall. The workforce will change shape, he argued, rather than simply shrink. "We need product managers who can vibe code and who can actually understand businesses," he said. "Software developers are of lower demand, but we need more data scientists."
The remarks echo a growing chorus of technology executives making similar predictions about the speed of AI-driven workforce transformation. WiseTech Global CEO Zubin Appoo declared earlier this year that "the era of manually writing code as the core act of engineering is over," as his company announced plans to cut up to 2,000 jobs across 40 countries - nearly a third of its global workforce. Cloudflare CEO Matthew Prince stated more bluntly that AI had made 1,100 jobs at his company obsolete, even as the company posted revenue growth of 34% year on year.
SAP's argument: the data advantage
Klein is quick to try to bolster his company’s position, saying SAP is structurally different from many of its software peers, arguing that the company's proprietary business data gives it an asset that generic AI models cannot easily replicate.
"I can't speak for every software company on the planet because I definitely feel for some of those niche vendors there is a huge disruption potential," he said, "but a few weeks ago I was sat with Dario Amodei, Anthropic's chief executive and his leadership team, and they said SAP has something they don't because we have AI that knows how to run companies."
SAP's ERP system - the software that manages core business operations for thousands of the world's largest companies - holds what Klein described as the "brain of the company": 7.3 million data fields covering pricing, logistics, finance, approvals and procurement. His argument is that AI agents can only deliver meaningful business outcomes when they have access to that kind of operational context, which SAP holds and large language model providers do not.
It is a version of the same argument being made across enterprise software, and so far the market has been sceptical. Still, at SAP's Sapphire conference in May, co-founder Hasso Plattner told Klein the company needed to "completely re-invent" itself - to forget everything product managers had learned and "think greenfield" about autonomous business operations.
The question remains, however, how long before AI starts amassing the kind of data that SAP has ringfenced at the moment?
What the data actually shows
Klein's prediction is stark, but the labour market evidence so far presents a more complicated picture - one HR professionals need to consider carefully.
Erika McEntarfer, a labour economist at Stanford and former Commissioner of the US Bureau of Labor Statistics, told HRD this month that the aggregate data does not yet show widespread AI-driven displacement. "It's not a story of mass displacements," she said. "What the aggregate trends are telling us is hiring has continued to happen apace." The US unemployment rate held at 4.3% in May 2026, with nonfarm payrolls rising by 172,000.
But the headline stability masks significant disruption at the edges. Stanford research has found that employment for young software developers fell nearly 20% by July 2025, and the entry door for early-career tech workers has narrowed sharply. The Federal Reserve Board found that employment growth among US programmers dropped roughly 50% after ChatGPT launched, falling from around 5% annually to near flat. As McEntarfer put it: researchers are currently debating whether software development jobs are growing as fast as they were before AI arrived, not whether they're disappearing altogether.
The broader displacement picture is also contested. A BCG microeconomic model published in April 2026 found that while 50-55% of US jobs will be reshaped by AI over the next two to three years, only around 12% face direct substitution - and BCG explicitly warned that companies cutting workforce beyond AI's actual ability to replace it would see productivity drop, institutional knowledge disappear, and critical talent walk away.
That warning is being tested in real time. As HRD has reported, a growing number of organisations that cut staff citing AI are now scrambling to rehire humans, as the technology proves less capable of independent operation than executives assumed. Forrester's J.P. Gownder has described an "AI washing" trend where companies attribute layoffs to automation that are primarily financially motivated - a distinction that carries increasing legal weight as state-level AI workforce legislation begins to take effect.
The HR implications
For HR leaders, Klein's statement is less important as a prediction and more important as a signal of intent - and a reminder that the gap between an executive's public forecast and a company's actual workforce planning can be considerable.
Klein himself acknowledged that the transition, if managed well, need not require net job losses. That conditional is doing a lot of work. Managing that transition - identifying which roles are genuinely at risk, which can be redesigned around AI capability, which require reskilling, and on what timeline - is precisely the workforce planning challenge that falls to HR, and MIT's Project Iceberg research has found that AI is already technically capable of performing work equivalent to nearly 12% of the US labour market, representing roughly $1.2 trillion in annual wages.
The vibe coding trend Klein describes does not simply eliminate developers. It changes what development expertise is worth. As BCG notes, roles being reshaped by AI typically require the same people to develop new skills rapidly - and those skills are not materialising fast enough to meet demand. ManpowerGroup's 2026 Talent Shortage Survey of nearly 40,000 employers across 41 countries found that AI skills have overtaken engineering and IT to become the hardest to fill globally for the first time, with 72% of employers reporting difficulty filling roles.
The workforce that can do what Klein describes - product managers who can vibe code, data scientists who understand business ontology, engineers who can build and govern AI agents - does not yet exist at the scale that SAP and its peers will need. Building it, or buying it, is an HR challenge. So is managing the anxiety of the 110,000 people who work at SAP today and heard their CEO say their function may cease to exist.
How that message is communicated matters more than the timeline attached to it. When WiseTech's executive chairman compared human labour to AI at a ratio of $100 to $2 before an investor audience, the fallout was immediate and severe. Klein, speaking more carefully about transition and redesign rather than pure replacement, appears to have absorbed that lesson. Whether the workforce hears the nuance or just the headline is a communication problem that sits squarely with HR.
