Add gender-disaggregated data to future releases
Hi Anthropic research team and community,
First — thank you for making this dataset public. It's one of the most substantive open resources on real-world AI adoption and economic impact.
A gap I keep running into: the dataset has excellent geographic granularity (country-level usage, per-capita rates, task breakdowns) but no demographic breakdown by gender.
This matters a lot for researchers studying the AI equity question. Every major digital adoption wave — mobile internet, broadband, smartphones — has shown a measurable gender gap, with the gap largest in lower-income countries. If AI is following the same pattern, we need data to see it. Right now, this dataset can't answer that question.
Concrete request: even a simple aggregate breakdown (e.g., % of usage by gender per country, suppressed for countries under a minimum user threshold to protect privacy) added to the existing facet structure would be transformative for gender and development economics research.
I'm specifically trying to study AI adoption in Central America (Costa Rica, Guatemala, Honduras, etc.) and the gender dimension is central to the policy implications. I suspect many researchers working on AI and the Global South face the same wall.
Would love to know:
If others are hitting this same limitation
Whether there are any plans to include demographic breakdowns in future releases
Whether Claude.ai even collects optional gender data at registration that could be aggregated
Happy to discuss methodology for doing this in a privacy-preserving way.
Thanks,
Elena