It’s now clear that AI is not solely a sector. It is an overlay on every sector. In 2025, roughly 40% of every dollar invested in startups on the Carta platform went to an AI company. In early 2026, that figure jumped to 54%. And there’s little reason to think AI’s share of total venture investment will stop growing anytime soon.
The dominance of AI is reshaping how founders build and compensate their teams. Smaller teams are proliferating. Average headcount at Series D fell 29% from its 2023 peak to 131 employees in 2025. At Series B, the average dropped from 53 to 45 over the same period. At seed, the median team size is now just four employees.
As team sizes shrink, companies have fewer workers to pay. This means more available capital for those employees that remain. As a result, median compensation levels have begun to grow. Across all job functions, the median size of initial equity grants issued to individual contributors at companies on Carta is up nearly 11% over the past two years, while the median salary for ICs is up 6.4%.
Nobody has benefitted more from this uplift in compensation than the technical employees who are making it possible. Among AI/ML engineers, the median initial equity grant ballooned by 31% between January 2024 and February 2026, nearly tripling the rate of growth in grant size across the broader population of employees. Over that same span, the median salary for AI/ML engineers rose by 9.1%.
The market for startup hiring and compensation has shifted in plenty of other ways since the public launch of ChatGPT in November 2022. Overall hiring has slowed. Job departures have become less common. Net headcount growth has flattened.
Exactly what this new world of AI-fueled startups will look like as the technology continues to evolve is still to be determined. But one fact has already been established: The old rules for building a startup no longer apply.
Key highlights
A new normal for headcount growth: Companies on Carta recorded net headcount growth of 5,652 employees during January 2026, in line with monthly figures seen over most of 2025. Following the rapid headcount growth of 2021 and 2022 and the headcount contraction that followed in 2023, the market seems to have settled into a new equilibrium.
The physical economy is hiring: Most sectors saw more new hires than job departures over the course of 2025. But headcount growth was strongest in some areas of the physical economy, including hardware, where there were 1.7 new hires for every departure last year. Healthtech, SaaS, and medical devices were among the other sectors with the most growth. For some startups in these spaces, the transition from research into build mode is underway.
Smaller startups are shelling out for AI talent: Compensation packages are growing for AI/ML engineers across companies of all types. But the biggest gains can be found at the smallest startups. At startups valued between $1 million and $10 million, the median size of initial equity grants issued to AI/ML engineers has increased by 64% over the past two years. At startups valued between $10 million and $25 million, initial grant size is up 52%.
Hiring & headcount

An ongoing recalibration in how startups build their teams continued throughout 2025 and into the early days of 2026. VC-backed companies on Carta made 26,030 new hires in January—typically the busiest month for startup hiring—which was the slowest January since 2018. Hiring has now declined in four straight Januaries, marking an overall 65% decline from the recent peak in January 2022.
The number of total departures from VC-backed companies on Carta has also been declining gradually for the past several years. Layoffs have consistently been less common than employees leaving their roles by choice.
Hiring figures for recent months may still shift, as employee records are typically logged after board approval of share grants, but for now, the trend is clear: Across both hires and departures, there’s less movement among startup employees today than there used to be.

From 2019 through most of 2022, companies on Carta consistently combined to hire thousands more employees than they lost each month. At the apogee of this trend, in January 2022, net headcount grew by more than 55,000 workers in a single month.
Then, things changed—a shift that aligns neatly with the public launch of ChatGPT, reflecting the influence of AI on this market evolution. Since the start of 2023, monthly net headcount on Carta has hugged zero, even dipping into negative territory on several occasions. In 2025, however, monthly headcount growth began to rebound, reaching its highest levels since 2022. In January 2026, companies on Carta made 5,652 more new hires than they saw employee departures.

For most of 2021 and 2022, startups were engaged in a hiring spree. The industry was flush with cash from a record-breaking run of VC fundraising, and many companies opted to deploy that capital into their human resources, expanding their employee base at a rapid pace. In many instances, headcount was the strategy.
Today, that model has been inverted. Startups are building leaner by design. Smaller engineering teams are shipping more with AI. And this shift in strategy is reflected in the data on headcount. In January 2026, VC-backed companies on Carta made 26,030 new hires against 20,378 departures, with hires outstripping departures by a rate of 1.3x. Compare that to January 2022, when new hires were 3.8x more common than departures.

In January 2026, companies on Carta saw 12,161 employees leave their roles voluntarily, while 8,217 were laid off. Both figures will likely shift as additional employee movement is logged on the platform. The broader trend, though, is unmistakable: Job departures among startup employees have been growing less frequent for the past three years.
Compared to the recent peak in June 2022, voluntary departures are down 40%. Layoffs, meanwhile, are down 57% from a high point in January 2023. However, both of these figures are still well above the typical levels seen prior to 2021, when near-zero interest rates still proliferated. The new normal for job departures may be much lower than it was three years ago, but it’s also significantly elevated from the pre-COVID, pre-AI era.

In 2025, the hardware sector saw more headcount growth than any other common startup industry, with a hire-to-departure ratio of 1.7. Not far behind were medical devices, healthtech, and SaaS, each of which saw 1.4x more new hires than departures last year. There seems to be a clear link between these industries: The sectors adding the most headcount today are where startups are either building physical things or creating software that can help enable the physical economy.
Most sectors have seen an increase in the ratio of hires to departures over the past two years, reflecting a recovery from the rock-bottom hiring environment of 2023. There are a few notable exceptions, however: The energy, biotech, and gaming industries have all continued to see declining hire-to-departure ratios. In each of the past two years, the gaming industry has actually lost headcount.
Salary & equity

From January 2024 through February 2026, the median salary at startups on Carta valued at $1 million or higher rose by 5.1%, while the median equity package increased by 13%. Both of these figures saw notable upward movement in late 2025 and early 2026, after having previously experienced declines earlier in 2025.
In terms of equity, this represents a sharp reversal: During 2023, the average size of equity packages plummeted across the startup industry, a drop that coincided with a steep decline in startup hiring. As headcount growth has begun to recover, the scale of equity compensation has bounced back, too.

Most of that recent growth in equity compensation has been concentrated in employees at the individual contributor and manager levels, rather than executives. From January 2024 through February 2026, both ICs and managers experienced double-digit growth in the size of their median equity packages. Executive compensation packages, meanwhile, grew by just 1.2% over the same span.
Recent growth in salaries has been more consistent across different job levels. But the gains have been most significant for ICs, where the median salary has grown by 6.4% over the past two-plus years. After previously increasing steadily for many months in a row, the median salary for executives actually declined slightly between September 2025 and February 2026.

Typical salaries and equity compensation packages surged in the back half of 2025 for companies of all sizes. In terms of equity, the biggest gains came at startups valued between $1 million and $100 million, where the median equity package is now 16.7% higher than it was in January 2024. But other cohorts aren’t far behind.
Growth in salaries has mostly been consistent over the past two years across startups of various valuations, with the median salary in each of the three cohorts shown here rising by between 5% and 6%. By a slim margin, startups in the $1 million to $100 million bucket have experienced the most combined growth in their overall compensation packages.

Startups on Carta pay the highest salaries for employees in the product organization, with a median of about $177,000 for new hires made during the first quarter of 2026. The next-best pay is found in engineering, where the median salary rose to $167,000 in Q1.
Other job functions are gaining ground. The median salary in legal roles increased by 5.6% over the past year, reaching $156,000 in H2 2025, and the median salary in accounting climbed by 6.8% over that same period. Interestingly, both legal and accounting tend to be knowledge-based functions, standing in contrast to more technical functions like product and engineering. At least for now, salaries in these knowledge-based roles have remained resilient in the age of AI.

For startup employees in legal and accounting roles, however, those gains in salary have been at least partially offset by declines in equity compensation. Legal, accounting, and data are the only three functions that saw the size of median initial equity grants decline from Q1 2025 to H2 2025, standing in sharp contrast to the significant gains seen in other functions.
This includes product, where the median initial equity grant grew by 12.4% over the past year, reaching 0.064% of total company equity. Median initial grant size in sales rose by 15.2% over the same period. Marketing, operations, people, and customer success functions also experienced double-digit gains in the typical size of initial grants.
AI/ML roles

It’s a good time to be looking for work if you’re an AI/ML engineer. From January 2024 to February 2026, the median salary for workers in AI/ML engineering roles increased by 9.1% (compared to 5.1% across all positions), while the median initial equity package shot up by 31% (versus 13% in the broader market). And this sample only includes startups valued between $1 million and $1 billion, meaning it excludes employees at the largest and most valuable AI startups, where compensation has likely been rising even higher.
These figures reflect the premium that VC-backed companies are willing to pay for top talent in the most sought-after positions, particularly in terms of equity. With strong AI systems one of the key valuation drivers for startups raising new funding, those startups are willing to provide commensurate equity upside to the employees responsible for building and maintaining those systems.

This competition for AI/ML engineering talent has been most pitched among smaller startups. From January 2024 to February 2026, the median equity package for AI/ML engineering roles rose by 64% for VC-backed companies valued between $1 million and $10 million, and by 52% for those valued between $10 million and $25 million.
For companies of this size, equity is frequently their most valuable currency. Often, these smaller startups lack the capital required to pay huge salaries, their equity pools are still largely untapped, and they still have substantial valuation runway. It can be “cheaper” for these companies to increase equity compensation than it is for their larger, more valuable peers.Of course, it’s no guarantee that these smaller startups will see their valuations continue to grow. But if they do, their equity grants could increase in value by 10x, 100x, or even more, turning those grants into lottery tickets with significant upside.

Over the past several years, the total number of new hires in AI and data roles made by companies on Carta has increased significantly, with a nearly fivefold increase taking place between Q4 2020 and Q4 2025. For the first few years of this period, the vast majority of hires in these roles were under the title of data engineer or data scientist. More recently, AI-specific roles have grown more common. Compared to a year ago, the frequency of new hires in AI engineering roles was up 150% in Q4 2025. The frequency of new hires in data engineering and data scientist roles, meanwhile, continues to decline.

Two years ago, only a tiny fraction of employees in AI and data roles on Carta had the job title of AI engineer. By the end of 2025, that figure had climbed to 17.3%, with another 11.8% of these employees in other AI-centric roles. Whether the nature of the work itself is changing, or just the job titles, there’s been a clear expansion over the past two years in the percentage of AI and data roles that are AI-specific.
The growth of these AI engineering and other AI roles has come at the expense of data scientists and data engineers, two job segments that have experienced significant contraction in recent quarters. Today, less than a quarter of employees in AI and data roles are described as data scientists, down from more than 50% at the start of 2020. Legacy data functions have not collapsed, but AI-specific roles are rapidly becoming the dominant part of the stack.
Company composition

Over the past four years, the typical number of employees at early-stage startups that successfully raise venture funding has mostly held steady. For instance, average employee count at startups that raised a seed round in 2025 stood at 5.8, compared to 6.1 for seed-stage startups in 2022. At later stages, there’s a clear trend of smaller teams. This is most apparent at Series D, where average employee count fell to 131.3 in 2025, down 29% from where it stood in 2022. In many cases, companies today are able to reach the same checkpoints with smaller teams, leveraging AI to help make up the difference.

There’s significant variance in the typical employee count at startups across different key industries at the time those companies reach the same fundraising thresholds. For example, the average medical devices startup that closed a seed Series C round during 2025 had about 33 employees. In SaaS, meanwhile, the average Series C startup had about 104 employees. In fintech, that number skyrocketed to nearly 214 employees.
Across most fundraising stages, the smallest teams can be found in medical devices and biotech. The largest team sizes differ a little more depending on the stage. At Series C and before, hardware startups tend to be near the upper end of the spectrum, reflecting the fact that physical products require physical teams. At Series D, though, the largest team sizes can be found in SaaS, healthtech, and fintech.

Compared to their non-AI peers, the founders of AI startups are holding on to slightly more of their equity as they progress through the fundraising timeline. At the seed stage, the median AI founding team retains 56.1% of their equity, compared to 56% for non-AI founding teams. But that gap quickly widens. By Series B, the median AI founding team retains 27.3% of their company’s equity, versus 21.8% for non-AI founders.
This difference in equity ownership tends to compound across stages because many AI-native companies are able to raise rounds with higher valuations and less dilution than their non-AI peers. Founders who got in early on AI are arriving at later stages with more ownership intact. Given the valuations some of those companies are able to command, a few more percentage points of ownership may ultimately be worth tens or even hundreds of millions of dollars.

At the time it raises a Series A, the median startup has set aside 15% of its fully diluted equity for its employee option pool. This number inches up gradually from one fundraising stage to the next, reaching 16.4% by Series C and 19.6% by Series E+.
But not every startup resides at the median—and there can be substantial variance in the size of employee option pools across the same stage. A Series A startup at the 10th percentile devotes just 8.3% of its fully diluted equity to the option pool, while a Series A startup at the 90th percentile has set aside 26.9% of its equity for employees. At earlier stages, the 90th percentile for option-pool size tends to be about 3x greater than the 10th percentile. At later stages, that ratio drops closer to 2x.
Methodology
Overall dataset
This report comes from thousands of Carta Total Compensation customers with over 900,000 current salary and equity data points used by CTC . Other metrics in the report, such as those that describe employee movement, derive from the aggregate pool of more than 1 million employees currently working for the 55,000 startups that use Carta to manage their cap tables. In addition, certain metropolitan area data comes from the U.S. Bureau of Labor Statistics.
The data presented above represents an aggregated, anonymized view into the compensation practices of these startups. Companies that have contractually requested that we not use their data in anonymized and aggregated studies are not included in this analysis.
This report represents a snapshot as of April 24, 2025. Historical data may change in future studies. New companies signing up for Carta’s services will increase the amount of data available for the report.
Salary & equity
All salaries presented in this report are expressed in U.S. dollars. Except where indicated, total payroll numbers do not include any variable compensation, such as bonuses or commissions, that may be given to employees.
All equity values presented in this report are expressed as a percentage of fully diluted company shares.
In the sections on salary and equity trends, changes over time reflect updates to Carta Total Compensation bands. These benchmarks are updated once per quarter to incorporate data from newly hired employees as well as any adjustments to current employee compensation.
DISCLOSURE: This communication is on behalf of eShares, Inc. dba Carta, Inc. ("Carta"). This communication is for informational purposes only, and contains general information only. Carta is not, by means of this communication, rendering accounting, business, financial, investment, legal, tax, or other professional advice or services. This publication is not a substitute for such professional advice or services nor should it be used as a basis for any decision or action that may affect your business or interests. Before making any decision or taking any action that may affect your business or interests, you should consult a qualified professional advisor. This communication is not intended as a recommendation, offer or solicitation for the purchase or sale of any security. Carta does not assume any liability for reliance on the information provided herein. ©2026 Carta. All rights reserved. Reproduction prohibited.





