A comprehensive study by GitHub, “AI in Software Development 2024”, shows that a large number of professional developers use AI tools, but not all companies actively support it. The gap is particularly stark in Germany. Respondents cited improved code security as one of the main benefits of AI.
Advertisement
Over 97 percent of developers surveyed say they already use AI tools to provide “technical support in the software development cycle”, but only a fraction of companies encourage or allow it. This attitude of companies towards AI tools clearly varies across the four countries studied: in the United States, 87 percent of companies explicitly support or allow AI, in India 79 percent, in Brazil 76 percent and at the bottom is Germany 59 percent. For companies that explicitly prohibit AI, the situation is the other way around: Germany leads here with 14 percent, while the others average less than 10 percent.
Not all companies – especially in Germany – support AI tools. Some even ban them.
(Image: GitHub)
GitHub concludes: “As developers look for workflow improvements, those responsible should also consider broader strategic goals and rules.”
Self-interest is low in Germany
Germany also lags behind in terms of how many developers use AI not only at work, but also in private. In the United States it is 82 percent, in Germany only 70 percent. After all, almost a third of local developers have not yet shown personal interest in AI.
Now almost all developers use AI at work, but only a few use it privately.
(Image: GitHub)
Adoption rates are lower for companies with complex toolchains than for companies with simple software chains. GitHub concludes: “This suggests that AI coding tools play a role in smoothing workflows and reducing toolchain complexity.” But there is also the interpretation that a complex toolchain makes the introduction of AI tools more difficult.
Safety promises one main advantage – shift left
GitHub Copilot increases productivity A study by the company from May Shown. The current survey is looking at other reasons why developers use AI. 90 percent of US developers reported a “perceived improvement in code quality when using AI coding tools.” This is followed by 81 percent in India, 61 percent in Brazil and 60 percent in Germany. However, only 15 percent of Germans believe that AI clearly reduces code quality, as other studies show. In Brazil it is also 19 percent, but in the US it is only 3 percent.
Many developers, especially in Germany and Brazil, doubt that AI improves code quality.
(Image: GitHub)
Developers are relatively unanimous in their belief that AI makes it easier to familiarize themselves with new languages ​​(60 to 71 percent), and they are regularly using it for testing environments (31 to 41 percent).
Almost all developers see a big benefit from AI assistants in the area of ​​code security. 41 percent in India, 40 percent in Brazil, 38 percent in the US and 30 percent in Germany saw a “significant improvement”. The value here may prove that German developers attach more importance to security from the very beginning. Overall, the makers of the study see great potential for a shift towards security during the development process.
Almost all respondents believe that AI improves the security of code. The creators of the study see a shift to the left here.
(Image: GitHub)
AI knowledge increases job prospects
Everyone also agrees on this: With AI knowledge, your own value in the job market somehow increases. When asked about the increased, “significant boost”, German developers are somewhat more cautious (43 percent) than their Indian colleagues (56 percent).
The study also asks what people do with the time they have gained. In Germany, “improving collaboration with colleagues” and “designing systems” come in first with 47 percent, followed by “trying out upcoming technologies” (45 percent).
2,000 people participated in the study, 500 per country. They come from a variety of companies with more than 1,000 employees (86 percent with no overlap). Mostly they are developers, a smaller number are also data scientists and software designers.
(Who)
