Saudi Arabia, which has designated 2026 as ‘The Year of Artificial Intelligence’, has put AI at the heart of its plans for economic transformation.
Billions of dollars are being committed to infrastructure, platforms, skills and new AI capabilities, while businesses and government organizations are moving quickly to put the technology to work.
However, as investment and adoption accelerate, organizations face a practical question: how do we know whether AI is actually making them more productive and creating sustainable economic value?
One of the clearest fields where we can begin to understand the real impact of AI is software development.
Globally, developers were among the earliest and most enthusiastic adopters of generative AI, making software development an important test of whether individual gains from AI translate into measurable organizational value.
It’s also a question that JetBrains, the company behind development tools such as IntelliJ IDEA, PyCharm and WebStorm, is taking an active role in assessing.
The company, which has been in the software development business for more than 26 years, works with a number of leading Saudi organizations and is expanding its engagement in the Kingdom, including participating in LEAP 2026.
“For Saudi Arabia, the return on AI investment will ultimately be measured in outcomes rather than adoption figures. In software development, those outcomes will include better products, faster and more reliable delivery, and a growing Saudi engineering workforce capable of using AI without outsourcing its judgement to it,” explains Nadia Rinsky, Head of Business Development for the MENA region at JetBrains.
In JetBrains' January 2026 AI Pulse survey of 11,000 developers worldwide, 90 percent said they regularly used AI for coding and development work.
However, no more than 13 percent were using it across the full software development lifecycle, including code review and release pipelines.
The gap between those figures points to an issue many technology organizations now face: helping a developer generate code faster is relatively easy to demonstrate, while establishing whether AI has improved software delivery at the team or organizational level is more difficult.
That is partly because generating code is becoming easier, while the work that follows it is not. The volume of code produced is a poor proxy for productivity.
AI-generated code still has to work within an existing architecture, pass review and testing, meet security requirements and remain maintainable long after it reaches production.
Developers also remain accountable for the software they approve, regardless of whether a human or an AI system produced the original code.
A more useful assessment looks at whether teams are releasing software faster, reducing failures and rework, improving reliability and allowing experienced engineers to spend more of their time on higher-value problems.
This has particular relevance for Saudi Arabia because its AI investment is being made alongside a major expansion of its domestic technology workforce.
The Kingdom's digital workforce grew from 150,000 people in 2018 to 381,000 in 2024, while women's participation in the sector has risen from 7 percent to 35 percent.
AI could significantly accelerate the development of this talent. Developers entering the profession can use it to understand unfamiliar code, learn frameworks and solve problems that previously required more time and support.
The important distinction will be not just becoming faster at producing code but also developing the engineering judgement required to decide whether that code is secure, scalable and appropriate.
That distinction becomes more important with AI agents. An agent can inspect a repository, change multiple files, run tests and revise its work, which shifts part of the developer's job from carrying out the work directly toward defining, supervising and reviewing it.
For organizations, this changes what needs to be measured. Once different teams are working with multiple assistants, models and agents, technology leaders need to understand which tools are improving delivery, what they cost, how their output affects quality and, where agents can take actions, what systems and data they can access.
As AI generates more software and takes on more complex tasks, visibility, verification and control become more important, not less.
Looking ahead, Saudi enterprises will need the freedom to choose between models and providers as the technology evolves, while retaining control over how those models access company data, interact with internal systems and operate within development environments.
The objective should not be to restrict experimentation, but to create the visibility and guardrails that allow successful experimentation to scale.
“As organizations move from AI pilots into production, the leaders will be those that can show where AI is creating value while maintaining confidence in the software they produce,” concludes Rinsky. “For Saudi Arabia, the opportunity is to turn the scale of today’s AI investment into sustainable capability and measurable outcomes.”
