More than two-thirds of Americans believe artificial intelligence will widen the gap between the rich and the poor, according to a new national poll. The finding underscores a growing public anxiety that AI is not a neutral force of progress but a tool increasingly calibrated to benefit those who already control capital, data, and infrastructure. This perception isn’t rooted in speculation—it’s a direct response to observable trends: automation displacing low-wage workers, tech giants amassing unprecedented profits from AI-driven efficiencies, and a venture capital frenzy that funnels billions into startups promising disruption while delivering concentrated returns.
Automation’s Uneven Toll
The fear isn’t abstract. From warehouse robots replacing human pickers to AI-powered customer service systems eliminating call center jobs, the labor market is already feeling the pressure. The poll shows that 68% of respondents expect AI to eliminate more jobs than it creates over the next decade, with low-income and service-sector workers bearing the brunt. These aren’t just blue-collar roles—entry-level white-collar positions in data entry, basic accounting, and administrative support are also at risk. What’s more, the jobs being automated are often the most accessible to workers without advanced degrees or specialized training, leaving fewer pathways into the middle class.
Meanwhile, the benefits of AI-driven productivity are flowing upward. Companies deploying AI report significant cost reductions and margin improvements, but those gains rarely translate into higher wages or broader hiring. Instead, they’re captured as shareholder returns or reinvested in further automation. This dynamic reinforces a familiar pattern: technological advancement increases aggregate wealth while concentrating it in fewer hands. The result is a system where productivity soars, but prosperity doesn’t trickle down.
Who Owns the Algorithm?
At the heart of the inequality concern is ownership. AI systems are built on massive datasets, proprietary models, and cloud infrastructure—assets overwhelmingly controlled by a handful of corporations. These firms don’t just develop AI; they monetize it through licensing, subscription services, and embedded features in existing products. The value generated by AI is thus privatized, while the societal costs—job displacement, retraining needs, strained public services—are socialized.
Consider the rise of generative AI. Tools like large language models can draft emails, write code, or generate marketing copy in seconds. For businesses, this means higher output with fewer human inputs. For workers, it means increased competition and downward pressure on wages. Yet the creators of these models—often PhDs and engineers at well-funded startups or tech giants—are rewarded with equity and high salaries, while the broader workforce sees little upside. The asymmetry is stark: those who build and own AI benefit disproportionately, while those whose labor is rendered obsolete are left to adapt on their own.
This isn’t a failure of innovation—it’s a feature of the current economic structure. AI doesn’t create inequality by itself, but it amplifies existing power imbalances. The technology is neutral in theory, but in practice, it operates within a system that prioritizes capital over labor, efficiency over equity, and private gain over public good.
The Myth of the Level Playing Field
Proponents of AI often argue that it democratizes access to tools and information, leveling the economic playing field. In reality, access remains deeply unequal. Small businesses, for instance, may use AI-powered analytics or chatbots, but they lack the data volume and computational resources to train custom models or compete with tech giants on innovation. The same algorithms that help a startup optimize ad spending also enable a multinational corporation to dominate search, advertising, and cloud markets.
Education and digital literacy further entrench the divide. While AI literacy is becoming a prerequisite for high-paying jobs, access to quality training remains limited. Online courses and bootcamps exist, but they’re often expensive or require time and bandwidth that low-income individuals can’t spare. The result is a two-tiered system: a skilled elite that leverages AI to advance their careers, and a majority that watches from the sidelines as the economy transforms around them.
Even when workers adapt—learning to use AI tools or transitioning to new roles—the pace of change outstrips institutional support. Retraining programs are underfunded, unemployment benefits are inadequate, and labor protections haven’t kept up with the rise of gig work and algorithmic management. The burden of adaptation falls on individuals, not systems.
This isn’t inevitable. Other countries have experimented with policies to distribute the gains of automation—universal basic income trials, robot taxes, and worker representation on AI oversight boards. In the U.S., however, policy responses have been fragmented and reactive. The focus remains on accelerating innovation, not ensuring its benefits are shared.
The poll’s findings reflect a public that isn’t anti-technology, but pro-fairness. Americans aren’t rejecting AI—they’re demanding a reckoning with its distributional consequences. The question isn’t whether AI will change the economy, but who gets to shape that change. Without deliberate intervention, the machine will keep running as designed: optimizing for efficiency, rewarding capital, and leaving labor behind.