Never mind.
Prominent artificial intelligence (AI) executives are walking back their dark predictions of a looming white-collar employment apocalypse.
OpenAI CEO Sam Altman admitted this week that his previous warnings about AI wiping out entry-level positions were wide of the mark. Speaking virtually at a Commonwealth Bank of Australia (CBA) conference in Sydney, Altman confessed his economic intuitions had been “just off.”
“I don’t think we’re going to have the kind of jobs apocalypse that some of the companies in our space advocate or talk about,” Altman said, adding he was “delighted to be wrong.”
The about-face also marks a stark departure from the doom-laden rhetoric that followed the 2022 launch of ChatGPT. Altman previously predicted AI would “replace most of the jobs people do today” and completely erase entire employment categories.
Last year, Anthropic CEO Dario Amodei warned that up to 50% of entry-level office roles could dissolve within five years, potentially driving global unemployment to 20%. Today, Amodei says AI could promote employment.
Change of heart or financial timing?
The reversals of Altman and Amodei come as their respective firms gear up for potentially blockbuster IPOs. Industry analysts suggest the sudden optimism from tech leaders may be strategic. OpenAI and Anthropic are currently angling for blockbuster public offerings and massive funding rounds. OpenAI is chasing a staggering $280 billion revenue target by 2030, while Anthropic is reportedly seeking a $30 billion funding injection. (On Thursday, Anthropic said it raised $65 billion in venture capital, giving it a valuation of about $965 billion to become the world’s most valuable AI startup.)
Meanwhile, recent economic data shows AI has had minimal impact on the job market despite recent layoffs. A May study by the Yale Budget Lab found no meaningful change in unemployment for workers in sectors highly exposed to AI. Experts note that technological capability is vastly outpacing actual workplace adoption. A recent Brookings Institution report highlighted that AI integration remains “costly and uneven,” failing to translate automatically into broad macroeconomic shifts.
The absence of an immediate AI jobs apocalypse doesn’t necessarily mean the underlying risks disappeared. It may simply mean we framed the disruption too narrowly, according to Serena Huang, an AI strategist and author.
Indeed, as concerns around job displacement grow and scrutiny intensifies, the conversation is being recalibrated in real time, said Lomit Patel, author of “Lean AI” and chief marketing officer of several startups. “What’s changing is not just the framing of AI’s impact, but the pressure on companies to manage perception alongside progress,” Patel said.
“The ‘AI will eliminate huge portions of the workforce overnight’ narrative was always too clean for how enterprise adoption actually works,” said Chae O’Brien, founder of Thought Bakery. “Fear is useful for accelerating market attention, but real organizational adoption is slower, messier, and far more behavioral than the frontier labs initially framed publicly. And if you look at what the major labs are doing this quarter, the strategy becomes much clearer. OpenAI and Anthropic both moved aggressively into forward-deployment and enterprise partnership structures. Long-term revenue resides in enterprise contracts, integration layers, operational embedding, and workflow dependency.”
If only most Americans agreed with Altman and Amodei.
“Public opinion research has made pretty clear that Americans feel quite negative about AI,” said Peter Wildeford, head of policy at the AI Policy Network, told Time magazine. “The AI industry, and in particular Sam Altman, has responded with an about-face… It’s hard to say whether they’ve actually changed their forecasts, or whether they’re just trying to change the narrative.”
The narrative shift also coincides with growing corporate anxiety over the exorbitant costs of implementing AI. Rather than saving companies money on human labor, advanced AI models are proving to be budget-draining liabilities.
Uber Technologies Inc. Chief Technology Officer Praveen Neppalli Naga recently admitted the ride-hailing giant burned through its entire 2026 budget for Anthropic’s Claude Code in just four months.
Bryan Catanzaro, vice president of applied deep learning at NVIDIA Corp., revealed that operating expenses are flipping traditional corporate math. “For my team, the cost of compute is far beyond the costs of the employees,” Catanzaro told Axios, suggesting AI could cost firms more than the humans they employ.
Microsoft Corp. has reportedly begun canceling licenses for its engineers to use rival software due to unsustainable expenses.
Despite the lack of a widespread systemic collapse, the tech sector itself is feeling a distinct pinch. Data from outplacement firm Challenger, Gray & Christmas reveals that AI spend and innovation have been cited as justification for nearly 50,000 job cuts this year.
Major firms continue to restructure; Meta Platforms Inc. recently eliminated 10% of its workforce, or 8,000 jobs, while financial software giant Intuit Inc. slashed 17% of its staff (3,000 people) to pivot toward AI. Google parent Alphabet Inc. and Amazon.com Inc. have announced rounds of layoffs as they push for greater AI adoption and efficiency.
While top executives now argue that the “human part” of employment will protect most white-collar workers, the financial reality remains complex. As workplace expert Andy Challenger noted, individual jobs may not be directly replaced by robots just yet, but corporate budgets are undeniably being diverted away from human roles to fund expensive AI infrastructure.
“The ‘AI jobs apocalypse’ framing was always the wrong lens. The real shift isn’t jobs disappearing,” Sahin Boydas, founder and CEO of Lovie.co, said. “It’s the fundamental unit of work changing. You no longer need 10 people to do what 10 people used to do. We’re entering the era of the solo founder: one person, the right tools, and a company that scales.”
“The shift in rhetoric around an imminent AI jobs apocalypse is colliding with the realities of enterprise deployment,” said Joseph Hoefer, chief AI officer at Monument Advocacy. “For two years, predictions of mass white-collar displacement dominated the conversation, but now, with overwhelming research itself showing the disruption hasn’t materialized at the predicted pace, the framing is shifting back toward augmentation, productivity gains, and gradual restructuring.”
Still, there’s real anxiety among younger workers. A recent survey by 11x found 80% of college seniors believe AI is already reducing entry-level opportunities.
Recent college grad Zachary Nelson acknowledges AI improves productivity, but at a cost. “With the experience from senior roles, junior roles get that needed training to benefit the company,” he said. “If that pipeline of traditional training goes away, what does that mean for the next generation of workers? That is my concern.”
“The priority now shouldn’t be fearmongering about mass unemployment, but ensuring companies use AI to augment talent, create new pathways into the workforce and help employees operate at a higher level,” said Prabhav Jain, CEO of 11x.
“Workforce change is still coming, particularly for repetitive tasks and junior knowledge work, but at a pace and shape that looks very different from the earlier warnings,” Hoefer said. “But I do think the cost of that earlier framing has been real for AI policy. Apocalyptic predictions raised the temperature in Washington and statehouses, hardened opposition, and made it harder to build durable consensus on issues that really do matter such as preemption, workforce transition, sector-specific deployment standards, etc. AI is likely to change almost every job faster than it fully eliminates them, and policy ought to be built for that reality.”
Added John K. Thompson, author of “The Path to AGI” and University of Michigan professor, “We will see increased productivity, and we will see roles automated away by AI, but we will not see wholesale jobs being destroyed by AI. I am AI optimist and an AI practitioner. There is much we can and will do with AI and if we do it well, we will elevate the elements that make us most human while automating those parts of work that are mundane and boring.”
Indeed, shifting the narrative from doom-laden rhetoric to a more optimistic outlook on employment might merely be a strategic move to appease investors and regulators rather than a purely data-driven change of heart.
Codestrap CEO Connor Deeks deemed the flip-flop a simple case of the PR machine “kicking into high gear because (OpenAI and Anthropic) finally realize that public sentiment has dramatically turned against them and their vision for an AI ‘utopia’ that none of us asked for.”
“This U-Turn is a thin marketing pivot away from what their true beliefs and pursuits are, and I imagine they U-turn once again after IPO,” Deeks said. “Thankfully, their predictions about AI’s impact haven’t materialized, at least in any measurable, beneficial way.”
