More Developers are Learning to Code Using AI tools – But That Doesn’t Mean They Trust Them

Despite this, developers remain pervasive in their scepticism, particularly regarding their trust in AI tools’ output. This division between AI’s transformative potential and developers’ cautious trust in its capabilities forms the crux of the evolving relationship between coders and their AI assistants.

AI tool usage in the UK is far below the global average

The developer community’s use of GenAI tools is growing, and their experiences are largely favourable. Our 2024 Developer Survey—which deep-dives into the sentiments and preferences of over 65,000 global developers, 3,200 of whom are based in the UK—found that the majority of respondents (76% of developers worldwide) are using AI tools in their development processes or plan to do so in the future.

Just over half (51%) of UK respondents in this year’s survey highlighted they are using AI tools somewhere in their development process, significantly up on last year’s figure (37%)—albeit far below the global average of 62%. Nonetheless, developers who are using the tools are generally happy with them. 62% of those surveyed in the UK see AI tools as very favourable or favourable tools for their workflow, just below the global average (72%).

However, 29% claimed to be ‘indifferent’ towards using AI tools as part of their development workflows, while 1 in 10 (10%) take a ‘completely unfavourable’ stance –  making the UK one of the least favourable countries to AI in Europe along with Poland (59%) and Germany (60%).  A further 36% of UK developers even stated that they are not currently using AI in their development process and do not intend to do so. Which, of course, begs the question – why?

UK developers split on the trustworthiness of AI tools

Developers may view AI tools sceptically due to concerns over reliability and integration challenges. 68% of UK developers cite a lack of trust in output, and 65% of UK developers cite a lack of codebase context as a top challenge with AI at work. These tools, while promising, often come with a steep learning curve and unpredictable results, making them more of a hindrance than a help in a high-stakes development environment.

Additionally, developers may feel that the opacity of AI decision-making processes undermines their control over project outcomes, leading to a preference for more transparent and manageable development tools. This combination of factors may contribute to a cautious stance towards AI, emphasising the need for solutions that better align with developers’ practical needs and preferences.

Interestingly, UK developers are split on the trustworthiness of the output from AI tools. The Developer Survey found that UK respondents deem the accuracy of AI tools as more untrustworthy (40%) than trustworthy (38%), with a further 22% taking a neutral stance.

UK leads the way with salaries

Developers in the UK may not place the greatest trust in GenAI tools, but their employers trust them a great deal – and aren’t afraid to put their money where their mouth is. UK median salary of every single job role listed in this year’s survey is higher than that of the EMEA average with Engineering Manager ($142,000 per annum), Blockchain Developers ($137,000 per annum) and Site Reliability Engineer ($127,000 per annum) the most lucrative roles.

This is also reflected across the board among individual coding languages. Developers using OCaml, Scala, Solidity, and Erlang top the list for commanding the highest salaries in the UK, with all of those listed in this year’s survey achieving higher averages than the EMEA median salary. This highlights that code development remains rewarding in the UK.

Humans + AI

Almost 40% of UK developers are still not fully sold on using AI tools for work – and trust and quality play a role in this. As Generative AI and LLMs continue to advance, the foundations of vetted, trusted, and accurate data will be central to how technology solutions are built. At the end of the day, content development is human-driven. GenAI can only respond with what has already been published – it won’t provide any data or insightful feedback on anything created after its last data ingestion.

Humans provide up-to-date information, identify patterns early on, and add signals around the social value of knowledge within a community, providing context to the content. While AI tools play a valuable role in coding, the value of the human community should not be underestimated.

Erin Yepis, Senior Analyst, Market Research and Insights Stack Overflow.