Atlas report
AI researchers in United States
Google (United States) is the largest, with 4 123 indexed researchers, barely ahead of Carnegie Mellon University at 2 905. No single place dominates here.
The short answer
157 641 researchers in the index have an institution in the United States, the 2nd largest count among 132 countries.
Per million inhabitants, the United States has 464 AI researchers in the index against 89 worldwide, 5.2 times the world figure.
Among people who started since 2024, the subject gaining fastest in the United States is Big Data and Digital Economy: its share of newcomers is 4.7 times what it was over the three previous years, with 407 people starting on it.
What stands out
With 4.2% of the world's people and 22% of its indexed AI researchers, the United States runs at 5.2 times its demographic weight.
Industry employs 17% of the United States's indexed AI researchers; worldwide the share is 8.9%, so the country runs at 1.9 times the average.
The subject where the United States weighs most against the rest of the world is Reinforcement Learning in Robotics: 2.4% of its subject assignments work on it, 1.3 times the world share of 1.9%.
Subjects
Topic Modeling is the largest subject in the United States, carried by 8.8% of its subject assignments.
Subject reports: Natural Language Processing Techniques, Multimodal Machine Learning Applications, Reinforcement Learning in Robotics, Generative Adversarial Networks and Image Synthesis.
| Subject | Researchers | Share here | Share worldwide |
|---|---|---|---|
| Topic Modeling | 25 979 | 8.8% | 7.9% |
| Natural Language Processing Techniques | 14 188 | 4.8% | 4.3% |
| Multimodal Machine Learning Applications | 11 429 | 3.9% | 3.9% |
| Advanced Neural Network Applications | 9 846 | 3.3% | 4.3% |
| Reinforcement Learning in Robotics | 7 086 | 2.4% | 1.9% |
| Generative Adversarial Networks and Image Synthesis | 7 016 | 2.4% | 2.6% |
| Teaching and Learning Programming | 1 216 | 0.4% | 0.2% |
| Big Data and Digital Economy | 537 | 0.2% | 0.1% |
Newcomers by year
The United States gained 37 746 new researchers since 2024, 1.9 times the 19 991 of the three years before; the world ratio is 1.8.
The 2025 intake for the United States was 0.8 times the average of the three previous years, against 0.8 worldwide.
| First publication | New in United States | Share of the world's newcomers |
|---|---|---|
| 2019 | 4 967 | 25% |
| 2020 | 5 357 | 26% |
| 2021 | 5 702 | 23% |
| 2022 | 7 515 | 22% |
| 2023 | 6 774 | 22% |
| 2024 | 7 694 | 26% |
| 2025 | 6 225 | 24% |
Institutions
6.2% of the country's researchers sit in three institutions, led by Google (United States).
Google (United States) is the largest, with 4 123 indexed researchers, barely ahead of Carnegie Mellon University at 2 905. No single place dominates here. Everyone else sits in places too small to name without pointing at individuals.
| Institution | Researchers | Share of the country |
|---|---|---|
| Google (United States) | 4 123 | 2.6% |
| Carnegie Mellon University | 2 905 | 1.8% |
| Georgia Institute of Technology | 2 717 | 1.7% |
Compared with its peers
Read against China and Germany, the two nearest countries by headcount, the differences are in the rates, not the totals. The gap that stands out is the share working in a company: 17% here against 7.5% for those two. Reports: AI researchers in China, AI researchers in Germany.
| United States | China | Germany | |
|---|---|---|---|
| Researchers | 157 641 | 238 132 | 32 275 |
| Per million inhabitants | 464 | 169 | 382 |
| In a company | 17% | 5.3% | 9.8% |
| Early career | 51% | 43% | 55% |
| Newcomers, recent against previous window | 1.9 | 2.5 | 1.1 |
Method
A country is where the institution sits, so someone who has moved and not yet published from the new address still counts at the old one. Counts move between two refreshes because the index grows every night. The method page sets out the sources and what each field leaves out.
Common questions
- How many AI researchers are there in United States?
- 157 641 researchers in the Founding Hires index have an institution in United States, 22% of all profiles with a known country, as of October 2026. The index counts people who published AI research, so this is a floor on the field, not a census.
- Where do you find AI researchers in United States?
- Google (United States), 4 123, Carnegie Mellon University, 2 905, Georgia Institute of Technology, 2 717. Their most common subject is Topic Modeling, 8.8% of the country's subject assignments.
- What is United States strongest in?
- Reinforcement Learning in Robotics, at 1.3 times the world share.
- Is the pool of AI researchers in United States growing?
- Yes, at the world's pace. Since 2024, 37 746 people in United States published for the first time, against 19 991 over the three years before, a ratio of 1.9 where the world sits at 1.8.
The names behind the numbers
Every count on this page is a list in the index.
Search it by subject and country to find who leads the work today, and set an alert on the ones about to come free, for $179 a month or 149 € in the euro area. Or hand the search over and pay only when someone joins.