AI’s Jobs Riddle Deepens as Layoffs, Growth Studies, and CEO U-Turns Collide

Mixed Headlines: Tech Layoffs Meet Surprising Job Growth Among AI Adopters

Artificial intelligence is already reshaping work, but the story it tells changes depending on which data you read and which CEO is speaking. Microsoft, Amazon, and Oracle have collectively shed tens of thousands of jobs since early 2025—even as they pour billions into AI infrastructure. Amazon alone cut about 30,000 positions between late 2025 and mid-2026. Yet many of the same leaders have publicly reversed course, calling AI an engine for future job creation after previously blaming it for workforce reductions.

New research deepens the confusion. A Ramp study of over 21,000 U.S. firms found that the heaviest AI spenders expanded overall staff by 10% and entry-level hiring by 12% over two years, while the bottom two-thirds of adopters saw zero headcount growth. Google’s internal research similarly portrays AI as a collaborative tool, not a wholesale job replacer. Meanwhile, a California Policy Lab analysis showed no statewide spike in unemployment insurance claims in AI-exposed roles since ChatGPT’s debut—but did find a significant jump in claims for college-educated workers in highly-exposed San Francisco positions.

The picture is further muddied by the hangover from pandemic overhiring. “The Big Tech companies definitely overhired during the pandemic and are now making the decisions to correct that overhiring,” noted Ara Kharazian, lead economist at Ramp. Many firms attribute layoffs to AI to appear forward-thinking—a practice researchers call “AI washing”—even when the cuts are largely a return to pre-pandemic staffing norms. In July 2026, nearly 200 economists and researchers, including Anthropic co-founder Jack Clark and former Google CEO Eric Schmidt, published a statement warning that AI could trigger “an unprecedented transformation of our economy, larger than the Industrial Revolution, but unfolding over a vastly shorter time frame,” urging policymakers to “act now.”

Decoding the Divergence: Why AI’s Net Effect Eludes Economists

Behind Big Tech’s Layoff Narratives

Amazon CEO Andy Jassy told employees roughly a year ago that AI would lead to a leaner workforce, then said in February that it could fuel job creation. Oracle, in a June SEC filing, explicitly tied recent layoffs to AI. These shifts make it nearly impossible for outsiders to isolate AI’s role from routine corporate restructuring. UCLA economist Till Von Wachter told Fortune it has been “notoriously hard to pin down” how much of the observed layoffs are really AI-driven. Until companies provide granular, auditable data, every claim is susceptible to spin.

The Ramp Data: Growth Among Intense Adopters, but Is It AI?

The Ramp study is the most granular expenditure-based look yet, but it comes with caveats. High-intensity AI adopters were disproportionately smaller, fast-growing companies that may have been expanding anyway. Moreover, as Stanford economist Erik Brynjolfsson pointed out, firms that adopt AI “may grow by gaining market share from non-adopters, so employment can rise among adopters even as exposed occupations shrink economy-wide.” In other words, net job growth in some pockets does not guarantee net job creation across the entire labor market.

Why Economists Are Sounding the Alarm Despite Positive Studies

The July 2026 statement from nearly 200 experts—including prominent figures who work inside the AI industry—reflects a belief that the current mixed data is a lagging indicator. They argue that AI’s displacement potential could materialize suddenly and at scale, just as the Industrial Revolution’s job destruction unfolded over decades but with a much steeper curve now. The California Policy Lab’s finding of elevated UI claims among college-educated, AI-exposed roles in San Francisco is an early warning of concentrated pain, even if it hasn’t yet shown up in national aggregates.

The CEO Pivot: From Job Killer to Creator and Back

The messaging whiplash isn’t just confusing—it has real consequences. An Amazon Employees for Climate Justice spokesperson told Fortune that staff feel “huge increased pressure” to finish tasks faster using AI, and that AI tools have raised output demands without making work easier. Meanwhile, OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei have both walked back earlier predictions of massive job elimination, now emphasizing that people will remain central. This pattern suggests that senior executives are struggling to balance investor expectations around AI efficiency with the need to maintain workforce morale and public trust.

What Workers, CEOs, and Policymakers Should Do Now

The uncertain outlook demands specific, near-term actions from each group affected. Here’s what the current evidence supports:

• Mid-career engineers and mid-level white-collar workers: Dave Clark, founder of AI logistics startup Auger, observed that mid-level staff may have the hardest time because they are less inclined to experiment with AI-driven workflows. Develop skills in AI-augmented problem-solving and system integration—the ability to understand how systems connect—rather than narrow domain expertise.
• Early-career and college-educated workers in tech hubs: The rise in UI claims among AI-exposed roles in San Francisco is a tangible signal. If you are in a highly exposed occupation, actively seek roles that involve verifying and refining AI outputs, which Clark and other executives say shows clear value.
• Corporate leaders: Avoid “AI washing” by transparently separating pandemic overhiring corrections from genuine AI-driven restructuring. A reputation for flip-flopping erodes employee trust and may invite regulatory scrutiny. Instead, publicly disclose AI’s impact on staffing plans with the same rigor used for financial results.
• Policymakers: The July 2026 economist statement’s call to “steer AI in a direction that complements humans” is urgent. Near-term steps could include funding transition programs for workers in concentrated AI-exposed roles, mandating better corporate disclosure on AI-driven layoffs, and investing in real-time labor market data systems that can track displacement before it becomes a broader crisis.
• Non-adopting firms: The Ramp study suggests that doing nothing carries competitive risk, as AI adopters are gaining market share. However, adoption should be paired with a clear workforce strategy—invest in retraining rather than using AI as a stealth layoff tool, because the reputational and regulatory risks of the latter are rising.