- 30.1% of US Python code is written by AI coding assistants
- Newer developers are even more likely to use AI
- Tech firms are also using more AI-generated code
A new research paper entitled “Who is using AI to code? Global diffusion and impact of generative AI” has found US software developers are the most intensive users of AI coding assistants globally.
By December 2024, artificial intelligence was believed to have generated nearly one in three (30.1%) Python functions by US developers on GitHub.
This puts US developers far ahead of their global counterparts in terms of AI usage, with countries like German (24.3%), France (23.2%), India (21.6%), Russia (15.4%) and China (11.7%) lagging behind.
US developers use AI coding assistants the most
The researchers also noted more experienced developers are less likely to use AI (28%) compared with newer GitHub users (41%) who might be more receptive to the platform’s latest additions.
Despite coming with huge productivity promises, AI doesn’t seem to have made such a great impact.
Moving to 30% AI-generated code has only correlated with a 2.4% increase in quarterly commits. The researchers place the economic value of AI-assisted coding in the US at anywhere between $9.6 billion and $96 billion annually, depending on the realistic productivity gains seen.
However, Daniotti et al noted at AI usage could be linked to greater experimentation, with a 2.2% increase in new libraries and a 3.5% increase in new library combinations observed, suggesting the tech could be helping developers expand into new programming areas.
The trend correlates with major tech firms like Google, Meta and Microsoft, which now admit that a large proportion (up to around a third) of their code, depending on project and use case, is generated by AI.
However, in the case of this study, the researchers admitted that the analysis focuses exclusively on open-source Python projects on GitHub, therefore the model effectively assumes that AI usage rates in Python are seen across other languages.
Still, they hope that quantified research could help AI sceptics make better-informed decisions about how they see themselves using AI and its effects on the labor market.
You might also like
- These are the best AI tools and best AI writers developers can use today
- Need an upgrade? Ask your boss for one of the best programming laptops
- Developers aren’t thrilled about having to add AI into everything they build, study shows