Software development in financial services enters new era as Gen AI supercharges productivity

A majority of software developers in financial services are using Gen AI daily in their work and over 90% report notable process improvements, including significant time savings. That is according to a study from DefineX, which surveyed over 600 practitioners and 70 executives in Turkey, the UK, and the Middle East.

The landscape of software development is undergoing one of its most significant changes in decades. Traditionally, progress was measured by how efficiently teams could write and test code.

But now, the rise of AI has now been shifting the focus from writing lines of code to the ability to express goals with clarity and precision. Experts say the industry is moving into a communication-first era where the instructions given to machines, rather than the code itself, are increasingly important for development.

Saving time with Gen AI

Current data shows that these tools have already become a regular part of everyday work for many in the industry. The survey found that in Turkey, more than 70% of workers use these tools every day, with most respondents saying it saves them from one to three hours weekly.

In the United Kingdom and the Middle East, roughly half of workers report the same level of daily use. While many developers use these tools to solve problems or draft documents, AI use is still mainly at the individual level rather than as part of a structured company plan.

Indeed, the benefits for individual workers are measurable. When it comes to the concrete benefits of AI, most respondents point to better productivity and enhanced process effectiveness.

On average, professionals are saving about three and a half hours every week, which is equal to roughly 10% of their total work time. These time savings help reduce fatigue from repetitive tasks and allow workers to get started on new projects more quickly.

Things are rapidly changing

Despite these gains, many company leaders in financial services remain cautious because it is difficult to see how these individual improvements translate into overall business success. In fact, many companies fail to really benefit from implementing AI tools for a variety of reasons, including poor implementation, lack of formal AI usage policies, legal and regulatory constraints, cost, privacy concerns, or other related issues.

As these tools take over more tasks, the roles of employees are also changing. Junior employees are finding that their traditional work, such as basic testing and routine coding, is the first to be handled by machines.

This change is forcing companies to rethink how new talent is trained and how they contribute. Meanwhile, senior developers and architects are shifting their focus toward checking and managing the work produced by Gen AI tools. This shift requires new skills and a different approach to training within the workforce.

Among the various barriers to further AI implementation, the top concern for executives has to do with legal regulations, which are especially tough in Europe, for example. As far as practitioners’ concerns, the majority (52%) said cost and potential privacy issues are restricting the use of AI in their companies. Again, the lack of unified rules for AI use in many companies has been a problem.

The future likely to bring a more structure

Looking ahead, the next two years are expected to be a period of more changes in how companies use these technologies. While many organizations are currently in a trial phase, experts predict that by 2026, the focus will shift toward creating official rules and systems to manage these tools at a larger scale.

For many financial institutions, the goal will be to move beyond simple tool usage and toward a future where AI is a core part of how software is developed. Success in this new environment will mean mastering how human ideas can be translated into precise instructions that machines can follow reliably.

“Across the industry, we see what can be called as intuitive coding: developers and analysts describe what they want, and AI generates code, documentation, or solutions in return,” stated the report’s authors.

“Yet something is often missing. Many organizations keep the generated output while discarding the specifications that created it. The path forward is to treat specifications as first-class assets and to build systematic capabilities around them. Organizations that succeed will not only generate code through AI but will also institutionalize how intent is expressed, validated, and reused across the entire lifecycle.”