Survey: The AI wave continues to grow on software development teams

Over the past two years, AI has become ubiquitous, appearing in everything from billboards to executive briefings. Last year, our inaugural developer survey revealed widespread interest in AI-powered coding tools among a small sample of U.S.-based developers, but we still had questions about how multi-disciplinary enterprise engineering teams were engaging with this technology.

This year, we broadened our survey by recruiting 2,000 total respondents with 500 respondents each from the U.S., Brazil, Germany, and India. While software engineers, developers, and programmers made up the majority of our respondents, we also included a small number of data scientists and software designers in our survey to get a fuller and more diverse view of AI’s impact. A consistent trend emerged: while our respondents say AI improves their experience in software development, their perceived usage in their companies remains slow.

More than 97% of respondents reported having used AI coding tools at work at some point, a finding consistent across all four countries. However, a smaller percentage said their companies actively encourage AI tool adoption or allow the use of AI tools, varying by region. The U.S. leads with 88% of respondents indicating at least some company support for AI use, while Germany is lowest at 59%. This highlights an opportunity for organizations to better support their developers’ interest in AI tools, considering local regulations.

Our survey respondents reported that AI helps them work more productively, using the saved time to design systems, collaborate more, and meet customer requirements better.

These findings suggest that individual AI usage isn’t enough. Organizations need to operationalize AI throughout the software development lifecycle to boost collaboration, creativity, and modernization.

AI doesn’t replace human jobs—it frees up time for human creativity. Now, let’s dive into the research.

Kyle Daigle
Chief Operating Officer // GitHub

Key survey findings

The generative AI wave in software development continues to grow. This year, we expanded our survey to 2,000 respondents—and almost everyone (upwards of 97%) reported that they have at some point used these tools both in and outside of work. (That’s not to say every one of their companies has sanctioned the use of these tools.)
While survey respondents say their organizations are welcoming AI, there’s still room for progress. Survey data indicates that a strong majority (59-88%) of respondents across all markets reported that their companies are either “actively encouraging” or “allowing” the use of these tools. To maximize the benefits of these tools, organizations should have a roadmap, a clear strategy, and policies in place to ensure wider adoption happens through building trust and driving measurable performance metrics.
Software development teams are recognizing more benefits with AI coding tools than previously reported. Some of these include building more secure software, improved code quality, better test case generation, and faster programming language adoption. This ultimately translated to time savings that they could use for more strategic tasks.

Methodology

This report draws on a survey conducted online by Wakefield Research on behalf of GitHub from February 26, 2024 through March 18, 2024 among 2,000 non-student, enterprise respondents in the U.S., Brazil, India, and Germany who are not managers and work at companies with 1,000-plus employees. This broke down to 500 respondents in each market and included people who identified their job titles as software engineers, developers, programmers, data scientists, and software designers. This broke down, by market, as follows:

This study was fielded online using a double opt-in research panel with around 86% of respondents coming from unique companies. This was a blinded survey; participants were never made aware of the sponsor, nor did they know the topic of the survey before agreeing to participate.

For the insights gained from this particular study, the chances are 95 in 100 that a survey result does not vary, plus or minus, by more than 4.4 percentage points within each market from the result that would be obtained if interviews had been conducted with all persons in each region represented by the sample.

💡 You can explore the full methodology behind this report with information on regions, industries, job titles, and more.

The growing AI wave in software development

Our survey data showed that nearly all of the survey participants reported using AI coding tools both outside of work or at work at some point. However, 17-27% of respondents indicated that they’ve only used AI tools at work, challenging the assumption that all developers are using AI outside of work.

GitHub has previously explored individual developers’ experimentation with AI, but not developers’ perspective into their organizations’ approaches to AI. So, we asked respondents to describe “companies’ approach towards the use of AI coding tools by software developers.”

In the chart below, we see 30-40% of those surveyed indicated their organizations actively encouraged and promoted the adoption of AI coding tools. An additional 29-49% of respondents across markets report that their organizations are allowing the use of these tools but offering limited encouragement. But there’s still room for organizations to actively join the AI wave.

💡The challenges behind organizational adoption of AI coding tools

At GitHub, we know that adopting AI coding tools at the enterprise level means that companies need policies that facilitate using these tools within workflows, while also considering factors like process changes, governance, and compliance.

While developers seek workflow improvements, leaders must also consider broader strategic goals and regulations. We see significant growth potential exists for developers and companies to leverage AI at scale as these technologies continue to evolve.

🔍 The difference we see between the reported usage of these tools and individual perceptions around organizational policies around these tools highlights the potential for broader AI integration and openness to using AI as part of a workflow.

Nearly half (48%) of respondents working in organizations that actively promote AI tools reported their toolchains were “simple” to use. In contrast, a significantly higher proportion (65%) of respondents from organizations with a neutral stance on AI use described their toolchains as complex. This suggests AI coding tools may play a role in streamlining workflows and reducing toolchain complexity for software development teams.

The reported benefits of AI coding tools

Our survey identified several key benefits that respondents associate with using AI coding tools in software development, including improvements in code quality, development efficiency, and streamlined workflows. Additionally, our survey suggests these tools are seen as facilitating upskilling and onboarding. By easing the transition to new programming languages and making it easier to understand existing codebases, these tools demonstrate material impacts for respondents in workplace settings.

Previous GitHub research has shown an up to 55% increase in productivity among developers who use GitHub Copilot, an AI coding tool. This led to the next natural question about how individual developers and teams will use the time saved with AI coding tools—which motivated us to ask the question directly to our survey respondents. But first, let’s explore the benefits respondents reported in our survey.

Building upon the identified developer-centric advantages, this section explores respondent’s perceptions around the broader benefits of AI coding tools for software development teams and organizations.

Improved code quality. Most respondents in the U.S. (90%) and India (81%), along with more than half in Brazil (61%) and Germany (60%), reported a perceived increase in code quality when using AI coding tools. This aligns with research we conducted with Accenture over the past year to analyze the impact of GitHub Copilot on enterprise engineering teams, and its impact on perceived code quality.

Easier to work with new programming languages, and understand existing codebases. A large portion (60-71%) of respondents reported that these tools make it “easy” to adopt a new programming language or understand an existing codebase. Notably, between 23-29% across countries reported AI coding tools made it “very easy” for respondents to adopt a new programming language or understand an existing codebase.

Test case generation. Overall, more than 98% of respondents reported their organizations have experimented with using AI coding tools to generate test cases. The majority of respondents reported their organizations use AI tools for test generation at least “sometimes.” That trend is most pervasive in the U.S. (reported by 92%) and the least pervasive in Germany (reported by 65%).

💡 The role of AI coding tools in generating test cases

At GitHub, we’ve seen the rise in AI adoption coincide with improvements in how AI coding tools cater to test case generation and security, too. These tools often build upon and complement traditional machine learning tools for test generation. This translates to faster test creation and improved coverage, even for legacy code that may have been overlooked before. However, it’s important to remember that AI-generated tests, just like code itself, require human review to ensure all potential scenarios are considered.

In our survey, respondents most commonly reported using the time they save with AI coding tools to design systems, collaborate, and learn. Specifically, 47% of respondents in the U.S. and Germany used this extra time for collaboration and system design. This continues a trend we first observed last year in a survey measuring AI’s impact on developer experience among U.S.-based developers, where respondents then reported that AI helped them focus on high-level tasks.

What are the expectations among those who have tried using AI at work?

Our survey indicates strong expectations among respondents that AI coding tools will significantly improve their ability to fulfill customer requirements. The majority of respondents (ranging from 61% in Germany to 73% in U.S.) expressed optimism about the potential of AI coding tools to moderately improve or significantly enhance their ability to meet customer requirements. This trend was consistent across various industries, suggesting a widespread expectation of benefits from generative AI.

Notably, the level of optimism among those surveyed seems to be associated with the company’s stance on AI use. Respondents who worked at organizations they reported as actively encouraging the use of AI were more likely to express confidence in the technology’s ability to drive customer satisfaction. This suggests a company that is supportive of the use of AI may help individuals maximize the potential application value of AI coding tools.

Respondents anticipate AI will enhance code security and development efficiency. There is near universal anticipation among survey respondents that AI coding tools will improve code security (99-100%). We can see the breakdown of the responses in the chart below, but notably we see the highest expectation of a significant improvement in India, with 41% of respondents expressing this view.

💡 AI’s potential in security: addressing the skills gap

The global shortage of security professionals is well documented. At GitHub, we understand that as an industry we can’t simply hire our way out of this problem. We’re aiming to address this challenge by reimagining shifting security left with developer-focused tools that make it easier to secure code from the start, by integrating into existing engineering workflows and suggesting solutions early in the development lifecycle. With recent AI-focused updates to our security products, we see strong promise for AI in helping advance these efforts.

Our survey reflects the importance of advancements like these. A large portion of our respondents, ranging from 59% in India to 67% in the U.S., said their security teams manually review code-based changes—and that is a bottleneck. AI tools such as Copilot Autofix in GitHub Advanced Security (GHAS) offer a promising solution to this problem by automating vulnerability identification and fixes.

Learn more >

Proficiency in AI coding tools is seen as a major asset by job seekers, according to our survey. Nearly all respondents (99-100%) believe this skill makes them more attractive candidates, underlining the growing importance of AI across various fields. Notably, a large portion (43% in Germany and 56% in India) believe this expertise significantly boosts their employability. However, it’s crucial to ensure this aligns with the specific AI coding skills employers are actively seeking.

💡 Knowing how to use AI now may give developers a career boost now. But AI is poised to become a standard part of software development moving forward.

Given AI coding tools such as GitHub Copilot have only been available for several years, this widespread expectation among our survey respondents about their positive impact on career advancement is notable. At GitHub, we anticipate that proficiency with using AI coding tools will soon become a standard requirement for software development roles.

Take this with you

Our research brings to light three crucial insights about the evolving landscape of software development:

Generative AI is rapidly transforming software development. Nearly all respondents in our survey have now tried AI coding tools, either personally or professionally (or both).
Respondents note multiple benefits when using AI coding tools. Collaboration and system design are the more strategic tasks where developers reinvest the time saved from using AI tools.
While survey respondents say their organizations are welcoming AI, there’s still room for progress. To realize the full potential of AI, companies should focus on fostering adoption through trust, clear guidelines, and measurable outcomes.

The potential of AI-driven software development is undeniable. By prioritizing a strategic approach that balances innovation, security, and organizational alignment, we can unlock its full potential—and this is an exciting time for engineering leaders to leverage these advancements and propel their engineering teams forward.

The post Survey: The AI wave continues to grow on software development teams appeared first on The GitHub Blog.

Source: Read More