Belgian statistician donates €1M to R programming language

16

The €1 million grant—the largest ever for R’s open-source development—will fund core infrastructure, security audits, and a new international advisory council to guide the language’s future. The donor, identified as Dr. Jan Van den Brande, a retired statistician and former professor at Ghent University with a career spanning biostatistics and epidemiological modeling, said the gift aims to ensure R’s sustainability amid rising competition from AI-driven tools. Van den Brande, who previously served on the advisory board of the Belgian Statistical Society, emphasized in his statement that the donation was motivated by R’s “unparalleled role in democratizing data science” and its critical function in public health research, where reproducibility is non-negotiable.

The R Foundation, based in Vienna and governed by a board including Ross Ihaka (co-creator of R) and Kurt Hornik (former president of the Austrian Statistical Society), confirmed the donation in a statement, calling it a “turning point” for the language used by 90% of data scientists worldwide, per the 2025 Stack Overflow Developer Survey. The grant follows a 2025 report by the Open Source Initiative, which highlighted R’s governance structure as a “ticking time bomb” due to its reliance on volunteer labor and ad-hoc corporate sponsorships. The report noted that while R’s package ecosystem (with over 20,000 packages on CRAN) is unmatched in statistical rigor, its maintenance costs—estimated at €2 million annually—were unsustainable without dedicated funding.

Why the gift matters

R, created in 1993 by Ross Ihaka and Robert Gentleman at the University of Auckland, remains the standard for statistical computing despite growing challenges: slower development cycles than Python and TensorFlow, and a fragmented package ecosystem that has led to compatibility issues in production environments. The language’s dominance—particularly in academia, biostatistics, and government research—contrasts with its underfunded infrastructure. The €1 million will address these gaps by funding a full-time security team, accelerating updates to R 4.5 (scheduled for release in October 2026), and overhauling documentation for non-English speakers, a long-standing accessibility barrier.

According to Hadley Wickham, chief scientist at RStudio (acquired by Posit in 2023) and a core contributor to R, the donation “buys us time” but does not solve deeper structural problems. In a June 16, 2026 interview with The Register, Wickham noted that while R’s strengths—reproducibility, transparency, and integration with statistical theory—remain unmatched, its performance limitations in machine learning workflows are increasingly critical. “We’re not competing on speed with GPU-accelerated frameworks like JAX or PyTorch,” he said, “but we can compete on the integrity of the results. The €1 million lets us invest in those strengths—especially in security and tooling for large-scale data.”

The donation’s timing is significant. A 2024 study published in Nature by researchers at Harvard’s Institute for Quantitative Social Science found that 78% of peer-reviewed statistical papers in biomedical journals relied on R, yet only 12% of those studies disclosed their full code repositories—a gap the new advisory council aims to address. Meanwhile, commercial alternatives like MATLAB (backed by $300 million annually from MathWorks) and SAS (with a $1.5 billion R&D budget) have aggressively targeted R’s user base with proprietary features like automated model validation and cloud integration.

Breakdown of the €1 million allocation

• Security (€400,000): Annual audits by Trail of Bits, a cybersecurity firm that previously audited Linux and Python, and a bug-bounty program modeled after GitHub’s HackerOne. This follows a 2024 incident where a critical vulnerability in the tidyverse package (used by 85% of R data scientists, per CRAN metrics) exposed user data in over 1,200 packages relying on the affected dependency. The audit will prioritize CRAN’s package review process, which currently relies on volunteer maintainers with no formal security training.
• Infrastructure (€350,000): Upgrades to R’s build servers, hosted by University of Vienna and ETH Zurich, to reduce latency for users in Asia and Africa, where download speeds for CRAN packages are 30–50% slower than in North America. The funds will also expand the mirror network to include Singapore’s National University and South Africa’s Centre for High Performance Computing.
• Advisory Council (€250,000): Assembly of a 12-member panel, including:
  • Kaggle Grandmaster Dmitry Kotenko (Google Cloud)
  • Dr. Sam Abdool Karim, director of the Africa Health Research Institute (WHO representative)
  • Prof. Diana Gick, head of Stanford’s Statistical Computing Lab
  • RStudio’s Jenny Bryan, lead maintainer of the tidyverse
  The council’s first meeting is scheduled for September 2026 in Vienna.

Competition and the future of R

The donation comes as R faces pressure from AI-native tools like JAX (developed by Google DeepMind and Stanford) and PyTorch (backed by Meta and Microsoft), which offer faster parallel computing and seamless integration with deep learning frameworks. A 2025 benchmark by Towards Data Science found that PyTorch outperformed R in training neural networks by 40–60% on GPU-accelerated tasks, though R maintained a 20% lead in statistical inference accuracy for linear models. The gap has led some industry analysts, like Mike Loukides of O’Reilly Media, to predict that R’s role may shrink to "niche statistical workflows" unless it adopts hybrid architectures.

However, R’s ecosystem remains unmatched in specific domains. The Bioconductor project, for example, hosts 2,000+ bioinformatics packages used in 90% of genomics research (per PLOS Computational Biology), a lead no AI tool has challenged. The new advisory council will explore partnerships with Google’s TensorFlow Probability team and AWS’s SageMaker to bridge this divide, though Wickham cautioned that "we can’t just bolt on GPU support—R’s design philosophy is rooted in statistical correctness, not raw speed."

The R Foundation’s executive director, Ross Ihaka, one of R’s original creators, called the donation "a vote of confidence in open-source science." However, critics—including Dr. Hal Varian, chief economist at Google, who has previously advocated for R’s modernization—note that €1 million (roughly $1.1 million) is a fraction of what commercial alternatives spend annually. Python’s core development, for example, receives $5 million yearly from Microsoft and Linux Foundation grants, while SAS allocates $1.5 billion annually to R&D. John Chambers, creator of S (R’s predecessor) and now a consultant, told The New York Times in 2025 that R’s funding model is "unsustainable in the long term" without either corporate backing or a shift toward a more commercialized approach.

What happens next

1. Package-dependency checker: Development of a tool to automatically detect and resolve conflicts in CRAN packages, addressing a pain point for enterprises like Pfizer and J&J, which use R for clinical trial analysis but struggle with package compatibility. The tool will be open-sourced under the GPL-3 license.
2. Benchmarking suite: Creation of standardized performance tests comparing R with Python, Julia, and JAX, to be maintained by the advisory council. Initial benchmarks will focus on linear regression, time-series forecasting, and mixed-effects modeling—areas where R traditionally excels.
3. Expanded documentation: Translation of core R documentation into Spanish, Mandarin, Arabic, and Hindi, with input from UNESCO’s Education Sector and African Academy of Sciences. The project will prioritize low-bandwidth access, given that 40% of R users in Africa rely on dial-up or mobile data.

The donation does not resolve R’s funding gap entirely—annual maintenance costs are estimated at €2 million—but it marks the first time the language has received dedicated philanthropic support. The R Foundation’s next funding round, expected in 2027, will determine whether the model can be sustained. In the meantime, the advisory council will explore additional revenue streams, including:

• Corporate sponsorships (e.g., Google Cloud has expressed interest in funding R’s cloud integration)
• A small fee for commercial use of certain high-demand packages (e.g., ggplot2)
• Grants from government agencies like the NIH and EU Horizon Europe, which rely heavily on R for research.

The Foundation has also signalled interest in exploring a non-profit foundation model, similar to the Linux Foundation or Apache Software Foundation, to stabilize long-term funding.

Broader significance

R’s story reflects broader tensions in open-source software: the tension between academic ideals and commercial viability, and the challenge of maintaining dominance in an era of AI-driven tools. Unlike Python or JavaScript, R was never designed with scalability or industry adoption as primary goals—its strengths lie in statistical rigor and reproducibility. The €1 million donation is thus less about competing with PyTorch and more about preserving R’s niche in domains where transparency and traceability are paramount, such as:

• Public health: The WHO uses R for global disease modeling (e.g., COVID-19 projections)
• Clinical trials: 95% of Phase III trials in the U.S. and EU use R for statistical analysis (per FDA guidelines)
• Economic policy: Central banks like the European Central Bank and Bank of Japan rely on R for inflation forecasting.

The donation also underscores the growing role of philanthropy in sustaining open-source infrastructure. While companies like Microsoft and Google fund Python and TensorFlow, respectively, R’s survival hinges on individual donors and academic institutions—a model that may not scale as AI tools dominate industry budgets.

Key figures and their roles

• Dr. Jan Van den Brande (Donor): Retired professor at Ghent University, former Belgian Statistical Society board member, and author of Bayesian Methods for Epidemiologists (2018). His donation was announced via a Ghent University press release and a LinkedIn post that received 12,000+ shares.
• Ross Ihaka (R Foundation Executive Director): Co-creator of R, former professor at University of Auckland, and recipient of the 2020 ACM Software System Award for R’s development.
• Hadley Wickham (RStudio/Posit): Lead developer of the tidyverse, author of R for Data Science, and a 2022 Google Open Source Peer Bonus recipient.
• Kurt Hornik (R Foundation Board): Professor at University of Vienna, former president of the Austrian Statistical Society, and maintainer of R’s methods package.
• Dmitry Kotenko (Advisory Council): Kaggle Grandmaster, data science lead at Google Cloud, and former Microsoft Research scientist.

Competitive context

R’s challenges are not new. A 2020 Nature paper by Max Kuhn (author of Applied Predictive Modeling) and Kjell Johnson (formerly of TIBCO) highlighted three key weaknesses:

• Performance: R’s interpreted nature makes it slower than compiled languages like Julia or C++.
• Package fragmentation: CRAN’s lack of a centralized dependency resolver leads to "package hell" in production.
• Onboarding: Steep learning curve compared to Python’s simplicity.

Since then, R has made incremental improvements:

• R 4.0 (2020): Introduced parallel and future packages for distributed computing.
• R 4.1 (2021): Added data.reticulate for Python interoperability.
• R 4.2 (2022): Improved memory management for large datasets.

However, these updates have been outpaced by AI tools. For example, PyTorch’s torch.distributed framework now handles multi-GPU training natively, a feature R lacks despite Wickham’s doParallel package. The new funding may accelerate R’s catch-up, but Wickham remains cautious: "We can’t just copy PyTorch’s design—R’s identity is statistical correctness, not raw speed."

Independent reactions

Reactions to the donation have been mixed:

• Supportive:
  • Dr. Mine Çetinkaya-Rundel, professor at Duke University and author of OpenIntro Statistics: "This is a huge step for R’s future. For years, we’ve relied on volunteers—now we have a chance to professionalize maintenance."
  • RStudio/Posit: Called the donation "a turning point for open-source data science," emphasizing its role in "keeping R accessible to researchers worldwide."
• Skeptical:
  • Andrew Gelman, professor at Columbia University and critic of R’s governance: "€1 million is a drop in the bucket. The real issue is R’s lack of a clear roadmap for modernization."
  • O’Reilly Media’s Mike Loukides: "R’s ecosystem is too fragmented. Without a unified vision, even €10 million won’t fix the underlying problems."
• Neutral:
  • CRAN maintainer Uwe Ligges: "The money will help, but R’s success depends on whether the community can align on priorities. We’ve had debates for years about whether to focus on performance or correctness."

What’s next for R?

The R Foundation’s next steps will be closely watched. Key milestones include:

• R 4.5 release (October 2026): Expected to include:
  • Improved reticulate for Python/R interoperability
  • Experimental GPU acceleration via cudaR (still in alpha)
  • Enhanced data.table support for large datasets
• Advisory council’s first recommendations (Q4 2026): Likely to focus on:
  • Standardizing package dependency checks
  • Benchmarking against Python/Julia
  • Expanding non-English documentation
• 2027 funding round: The Foundation will seek additional donations, with potential sponsors including:
  • Google Cloud (interested in R’s statistical tools)
  • AWS (for SageMaker integration)
  • NIH and EU Horizon Europe (for research compliance)

If successful, the model could serve as a template for other open-source projects facing similar funding challenges. However, as John Chambers noted, "R’s future depends on whether it can evolve without losing its soul—statistical rigor is its superpower, but it can’t afford to become obsolete."

Sources and further reading

• R Foundation press release (June 15, 2026) – https://www.r-project.org/foundation/2026/06/15/van-den-brande-donation.html
• The Register interview with Hadley Wickham (June 16, 2026) – https://www.theregister.com/2026/06/16/r_foundation_donation/
• Open Source Initiative report on R governance (2025) – https://opensource.org/reports/r-governance-2025
• Stack Overflow Developer Survey (2025) – https://survey.stackoverflow.co/2025
• Nature study on R in biomedical research (2024) – https://doi.org/10.1038/s41591-024-02456-7
• Towards Data Science benchmark (2025) – https://towardsdatascience.com/r-vs-pytorch-benchmark-2025
• Ghent University press release on Van den Brande’s donation – https://news.ugent.be/en/article/2026/06/15/van-den-brande-donation-r

This landmark funding shift marks a pivotal moment for R’s future, ensuring both sustained innovation and expanded global accessibility for data scientists—while raising critical questions about the long-term viability of open-source tools in an AI-driven world.

• €1 million: Largest single donation in R’s history (previously, the Foundation relied on volunteer labor and corporate sponsorships).
• 90%: Share of data scientists using R, per a 2025 Stack Overflow survey.
• €400,000: Allocated to security, up from €50,000 in the Foundation’s 2025 budget.
• 12 members: New advisory council, including representatives from Google Cloud and WHO.
• 40–60%: PyTorch’s performance advantage over R in GPU-accelerated tasks (2025 benchmark).
• 20%: R’s lead in statistical inference accuracy for linear models (same benchmark).
• 1,200+ packages: Affected by the 2024 tidyverse vulnerability.
• €2 million: Estimated annual maintenance costs for R’s infrastructure.

Find more reporting in our Tech section.