AI Jobs and Business Growth: What the Data Actually Shows

The AI Jobs Debate That Refuses to Settle

The conversation around AI jobs and business growth has never been more contested — or more consequential. Through May of this year, companies announced close to 90,000 job cuts directly tied to AI, according to tracking data cited by TechCrunch. Some projections suggest that up to 15% of U.S. jobs could be eliminated by AI over the next five years. For developers, IT decision makers, and policy professionals already navigating rapid digital transformation, these numbers are hard to ignore. And for a generation of graduates wondering whether anyone will be hiring when they finish their degrees, the promises from the tech industry that AI will "create new jobs" ring increasingly hollow.

But a significant new report from enterprise AI spend tracker Ramp and workforce analytics firm Revelio Labs — which collectively monitor data from nearly 22,000 companies — complicates the doomsday narrative in ways that deserve serious attention. The report doesn't claim AI universally creates jobs. What it does argue, with data to back it up, is that the relationship between AI investment and headcount is far more nuanced than the headlines suggest. For IT leaders and entrepreneurs deciding whether to commit to sustained AI adoption, the implications are significant.

What High-Intensity AI Adoption Actually Does to Headcount

The Ramp/Revelio Labs report introduces a useful concept: the "high-intensity adopter." These are firms spending an average of at least $30 per employee per month on AI tools within the first three months of adoption. The finding? Those companies saw headcount increase by 10.2%. That's not a rounding error. It's a meaningful growth signal that cuts against the dominant narrative of AI as a purely substitutive technology — one that replaces workers rather than enabling companies to hire more of them.

The headcount growth wasn't isolated to a single department, either. It spanned engineering, sales, administration, customer service, finance, marketing, and scientific roles. The strongest growth was in the information sector — software, internet, media, and tech-adjacent firms — which is unsurprising given those industries' structural readiness for AI integration. These are the same sectors that have been building out data infrastructure, investing in cloud computing, and cultivating the engineering talent pipelines that make AI adoption operationally viable.

Perhaps most striking for the policy debate: entry-level headcount in these tech-forward firms actually rose by 12%. This directly contradicts research from Goldman Sachs, which found that AI has already erased approximately 16,000 net jobs per month over the past year, with Gen Z and entry-level workers bearing the brunt. Both datasets can be true simultaneously — and that's precisely what makes the AI jobs debate so difficult to resolve with any single study.

AI as a Tool for Firm Expansion, Not Just Labour Substitution

The report's most important conceptual contribution may be its reframing of how AI creates value in certain firms. Rather than simply cutting costs by replacing workers, AI adoption in high-growth tech firms appears to lower the cost of core output — writing and debugging code, building internal tools, producing technical documentation, supporting product development — which in turn raises the return on expanding the whole business. When it costs less to build, you can afford to build more. And building more means hiring more.

This is a fundamentally different economic logic from the one underpinning most AI-job-loss fears. The standard narrative assumes a fixed pie: AI takes tasks, workers lose jobs. The Ramp/Revelio data suggests that in certain contexts, AI grows the pie — enabling firms to take on more projects, serve more customers, and scale operations without hitting the traditional marginal cost walls that cap growth.

For developers and IT leaders, this framing is worth internalising. If your organisation is using AI purely as a cost-cutting mechanism — replacing contractors, reducing support headcount, automating repetitive workflows without reinvesting the savings — you may be capturing only a fraction of the available value. The firms seeing the strongest headcount growth appear to be those treating AI as a platform for scaling ambition, not just trimming overhead.

This is also relevant from a digital sovereignty and AI regulation perspective, particularly in Europe. Policymakers considering how to regulate AI's impact on labour markets should be cautious about treating all AI adoption as equivalent. The context in which AI is deployed — and whether firms reinvest productivity gains — matters enormously for social outcomes. Research from the OECD on AI and employment has consistently highlighted that policy frameworks need to account for heterogeneous adoption patterns across firm types and sectors.

Why the Positive Headlines Come With Uncomfortable Caveats

The report's authors are admirably candid about its limitations. The dataset skews heavily toward tech-forward, knowledge-work firms — many of which may have venture capital backing and were already on high-growth trajectories before committing to AI. This makes it genuinely difficult to isolate whether AI investment is causing headcount growth or whether it's simply co-occurring at firms that were going to hire aggressively anyway.

This is a classic problem in technology adoption research, and it's one that honest analysts on both sides of the AI debate should acknowledge. Correlation between AI spend and headcount growth doesn't establish causation. The firms most likely to invest heavily in AI tools are also the firms most likely to have strong balance sheets, experienced engineering teams, and management capability — all of which independently predict growth.

The Goldman Sachs data, meanwhile, reflects a broader cross-section of the economy — including sectors like retail, logistics, customer service, and administrative functions where AI is far more likely to function as a substitution tool than an expansion enabler. When you average across the whole labour market, the job losses are real. The Ramp/Revelio findings don't negate that. They describe a specific subset of companies operating in specific conditions.

Critically, the report also finds that firms which bought subscriptions and ran pilots but did not go on to make sustained, high-intensity AI investments saw no meaningful gains in headcount. The "dabbler" approach — the AI equivalent of buying a gym membership and going twice — does not appear to produce the same organisational lift as committed, embedded adoption. This is a finding that small business owners and entrepreneurs considering AI strategies should take seriously.

How AI Adoption Is Widening the Gap Between Well-Resourced and Struggling Firms

Perhaps the most important policy-relevant finding in the report is what it implies about the distribution of AI's benefits. If sustained, high-intensity AI adoption requires capital, technical staff, founder networks, and management bandwidth — and if only those firms see meaningful headcount and productivity gains — then AI risks becoming another dimension along which resource-rich firms pull further ahead of everyone else.

The report's authors put it plainly: "Firms without those channels may fall behind." This is a concern with significant implications for European SMEs and startups that lack the infrastructure of Silicon Valley-scale organisations. European tech policy has rightly emphasised digital sovereignty and equitable access to technology — but access to AI tools and access to the organisational capacity to deploy them meaningfully are very different things.

For IT decision makers and policy professionals in Europe, this suggests two parallel imperatives. First, encouraging sustained AI adoption rather than superficial experimentation — through training programmes, subsidised infrastructure, and open-source tooling that reduces the capital barrier. Second, designing AI regulation that doesn't inadvertently widen the advantage of already-dominant firms by imposing compliance costs that only well-resourced organisations can absorb. The EU AI Act and GDPR compliance frameworks both have significant implications here, as smaller firms may find the regulatory overhead disproportionate relative to their AI investment capacity.