Starting a business with AI no longer requires venture capital; a $200‑per‑month subscription and a solid idea are sufficient.
U.S. business‑application data increasingly supports this trend.
Since early 2025, the number of new business applications has risen sharply, closely mirroring the rapid advances in generative AI and the emergence of agentic coding tools.
These applications are not for the next major AI startup; they reflect solo entrepreneurs, freelancers, and side‑hustlers who can now bring their ideas to life with affordable AI tools.
Growth of Single‑Person Enterprises Aligns with Advancing AI Technologies
The U.S. Census Bureau monitors monthly applications for new businesses and categorizes them into two groups.
- “High‑propensity” applications are those most likely to eventually hire employees (dark blue area) – for example, a new restaurant, a small manufacturing firm, or a professional services firm. These are the businesses traditionally used as a barometer of entrepreneurial activity by economists.
- “All other” applications encompass everything else (light blue area): primarily one‑person entities such as sole proprietors, single‑member LLCs, independent contractors, and e‑bay sellers, as well as businesses that do not indicate hiring plans for the coming year or planned wages. They may or may not add employees beyond the founder, but they still represent genuine economic activity.
There have been two notable shifts in application levels in recent years.
The first shift occurred during the COVID pandemic, when a wave of new entrepreneurs emerged as a result of pandemic‑related disruption, affecting both high‑propensity and all‑other groups.
The second shift began in early 2025, and the surge in business formation has been driven almost entirely by one‑person companies; their applications have risen by more than 20%, while high‑propensity applications remain nearly flat.
Chart 1: Business applications have risen almost exclusively due to one‑person enterprises
What prompted this change?
We attribute the timing to AI, specifically the step‑change in generative AI capability that emerged with agentic coding tools in early 2025.
These tools go beyond simple code autocompletion; they can autonomously perform multi‑step tasks such as building websites, drafting product descriptions, handling customer emails, and creating marketing content—tasks that previously required hiring personnel.
Sectors with Higher AI Adoption Show Greater Business Application Growth
To test this hypothesis, we segmented the application data by sector and correlated each sector with its AI adoption rate.
The result is that since February 2025, nearly half of the monthly increase in “all other” business applications originates from high AI‑adoption sectors (green area).
These sectors—technology, finance, professional services, and others—allow AI‑augmented solo operators to accomplish work that formerly required a small team.
Notably, similar results are observed when aggregating sectors by AI exposure—a metric indicating the proportion of tasks in each occupation that can be completed at least 50% faster by an LLM alone or with basic scaffolding, without sacrificing quality.
Chart 2: High AI‑adoption sectors account for half of the monthly growth in solo business applications
There is an important caveat, however.
Previously, most of the growth stemmed from medium‑adoption sectors (dark blue area), largely driven by the retail sector, which is likely unrelated to AI.
The American Rescue Plan Act reduced the IRS reporting threshold for platforms such as eBay, Etsy, and Venmo from $20,000 in 2022 to $2,500 by 2025, prompting many casual sellers to formalize as businesses. The One Big Beautiful Bill Act (OBBBA) later reversed this change in mid‑2025, potentially explaining the recent moderation in that tier.
Although the increase is real, it is likely a tax‑reporting artifact rather than an AI signal.
Nevertheless, monthly business applications in high‑adoption sectors have been steadily rising since February 2025.
High AI‑Adoption Sectors Have Been the Most Productive Since 2005
Beyond being an interesting fact about AI’s growth, this matters for the economy because sectors with the highest AI adoption rates are also the most productive in the U.S., by a wide margin, and over a long time horizon.
Over the past 20 years, productivity—measured as real gross value added per employee—in high‑adoption sectors has grown at an annual rate of 2.2% (green bar), compared with 1.6% in medium‑adoption sectors (dark blue bar) and a negative 0.1% in low‑adoption sectors (light blue bar).
Although the difference between 2.2% and 1.6% may appear modest, over 20 years it translates to 56% growth for high‑adoption sectors versus 39% for medium‑adoption sectors.
Chart 3: High AI‑Adoption Sectors Have Averaged 2.2% Annual Productivity Growth Since 2005
These are also the sectors where AI drives new business formation, as they offer the strongest economic case for AI assistance and where automation gains compound over time.
When a new business emerges in a high‑productivity sector, it contributes disproportionately to economic output relative to its workforce. A solo consultant leveraging AI to match a small team’s output exemplifies the productivity story that later appears in GDP data.
AI Is Reducing the Cost of Starting a Business (Early Stage)
However, we are still at the early stages.
Agentic coding has been widely available for barely over a year. The full suite of AI tools that will become routine for solo entrepreneurs in 2027 or 2028 does not yet exist.
Nevertheless, the data indicates that AI is unlocking the marginal entrepreneur.
These are not hot new startups; they are individuals whose business ideas were viable in concept but not in practice, because execution required one or two employees—costly, legally complex, and operationally demanding. GenAI dramatically lowers that barrier.
Because these new businesses cluster in the economy’s highest‑productivity sectors, their ripple effects could be significant. More competitors in high‑productivity fields foster greater innovation, higher output per worker, and—if history is any guide—a real boost to U.S. productivity growth over time.
If the trend persists, it adds an additional productivity channel for AI that may have been overlooked. It will not be limited to large companies automating workflows, but will also enable hundreds of thousands of small businesses that otherwise would not exist.