Identifying the best global experts in artificial intelligence requires choosing a measurement criterion. Media notoriety, number of scientific citations, control of strategic resources (chips, data, capital): each framework produces a different ranking. This article compares these approaches to understand who truly influences the development of AI in 2026.
Scientific Citations and h-index: The Bibliometric Framework for AI Researchers
Most public rankings mix executives, popularizers, and researchers. Lists based on bibliometric metrics (cumulative citations on Google Scholar, h-index) paint a tighter picture.
TrexoMedia ranks in 2026 Yoshua Bengio, Geoffrey Hinton, Yann LeCun, Fei-Fei Li, and Andrew Ng among the most cited AI researchers, explicitly based on recent research data. These five names appear in nearly all international scientific rankings.
Hinton and LeCun share the 2018 Turing Award with Bengio for their foundational work on deep learning. Hinton received the Nobel Prize in Physics in 2024, while Demis Hassabis received the Chemistry Prize the same year. These distinctions serve as objective markers that social media notoriety cannot replace.
A broader overview, including the strategic and economic dimension, is provided by AI experts on Le Comparatif, which crosses several evaluation criteria.
AI Executives and Investors: Expertise Measured by Resource Control
AI Magazine (BizClik) publishes a Top 100 AI Leaders in 2026, where the top ten are mostly CEOs or CTOs of large tech companies. This type of ranking is based on a different premise: whoever controls the budget, chips, and data controls the trajectory of AI.
| Profile | Main Criterion | Examples 2026 | Limitation |
|---|---|---|---|
| Academic Researcher | Citations, h-index, scientific awards | Yoshua Bengio, Geoffrey Hinton, Fei-Fei Li | Does not measure industrial influence |
| AI Lab Executive | Valuation, fundraising, deployed products | Sam Altman (OpenAI), Dario Amodei (Anthropic), Demis Hassabis (Google DeepMind) | Confuses economic power with technical expertise |
| Infrastructure Provider | GPU market share, data center revenue | Jensen Huang (Nvidia) | Hardware expertise, not algorithmic |
| Popularizer or Influencer | Audience, social media engagement | Andrew Ng, some content creators | Popularity without direct link to research |
Sam Altman is preparing for an IPO of OpenAI targeted for 2026, with a valuation that could approach historic highs for the sector. Dario Amodei, CEO of Anthropic, predicts that AI will surpass humans in nearly all cognitive tasks by 2027. These positions shape public debate as much as any scientific article.
Jensen Huang occupies a unique position: Nvidia provides the computing power without which no large model can be trained. His influence is infrastructural, not algorithmic, but it conditions everything else.
AI Expertise in France: Researchers, Entrepreneurs, and Trainers
France is home to a pool of skills recognized internationally. Yann LeCun, Turing Award winner and AI scientific director at Meta, remains the most cited figure. However, the French landscape is not limited to a single name.
- Luc Julia, co-creator of Siri and former CTO of Renault, represents the industrial side of French AI expertise, focused on concrete deployment in companies.
- Mistral AI, founded in Paris, has raised significant funds and is developing language models that compete with American offerings on certain benchmarks.
- The training ecosystem is structured around experts like Alexandre Kantjas or Julien Simon, who make machine learning and generative AI skills accessible to non-engineer profiles.
France already had over 166,000 job offers related to artificial intelligence recently, a sign that the demand for skills far exceeds the circle of researchers.
Criteria for Selecting AI Experts: What Rankings Don’t Tell
No single ranking captures the reality of expertise in artificial intelligence. Bibliometric metrics ignore industrial impact. Executive lists confuse decision-making power with technical mastery. Influencer rankings measure audience, not rigor.
A criterion rarely specified in rankings is the reproducibility of published works. A researcher whose results are replicated by other teams has a verifiable scientific impact, whereas an executive can influence the market without ever publishing a line of code.
- To evaluate a researcher: citations, h-index, scientific awards, reproducibility of results.
- To evaluate an executive: deployed products, user base, documented development strategy.
- To evaluate a trainer or popularizer: pedagogical quality, verifiable feedback from learners, updating content in response to the rapid evolution of tools.
The global AI market surpassed $500 billion in 2024. At this scale, the very notion of an expert fragments: no one masters both fundamental research in deep learning, corporate strategy around generative AI, regulation, and training. The most relevant profiles depend on the question posed, not a universal ranking.
The takeaway: bibliometric rankings and economic rankings almost never designate the same individuals. Choosing your framework for understanding is already choosing your answer.