Artificial intelligence is profoundly transforming the way software is developed in the business world. Its impact is already evident in day-to-day tasks such as code generation, refactoring, technical documentation, test automation and error analysis. However, to limit its value to a mere question of productivity would be to scratch only the surface. The real opportunity lies in how AI can help build applications that are more robust, more secure and better aligned with the demands for quality, scalability and regulatory compliance that characterise organisations such as Telefónica.
For years, the developer’s role has been shaped by the need to balance speed of delivery, technical soundness and security. Today, with the emergence of generative AI and intelligent tools applied to the development cycle, that balance is entering a new phase. Many repetitive or low-value-added tasks can be significantly accelerated: from proposing initial code structures to creating unit tests, reviewing queries, drafting documentation or the preliminary identification of incidents. This frees up time for developers to focus on what truly adds strategic value: architecture, business logic, service resilience, performance and decision-making based on technical judgement.
How artificial intelligence is transforming the day-to-day work of developers in enterprise environments
Within Telefónica, artificial intelligence no longer acts solely as an assistant for specific tasks, but as a capability that supports virtually the entire development process. Its impact is felt from the earliest stages – such as refining the specification and designing the architecture – right through to implementing the solution and designing tests. This is profoundly changing the way developers work; they are spending less and less time writing code manually and more time directing, validating and ensuring the quality of what is built with the help of AI.
In this new landscape, the developer’s value lies in activities involving greater technical responsibility: correctly understanding the requirement, translating it into a well-designed solution, validating that the generated implementation meets the expected requirements, and ensuring its proper integration within the technological ecosystem. Testing work is also becoming particularly important, especially integration testing, as it is key to verifying that the various components behave consistently, securely and efficiently in complex and highly connected environments.
Far from disappearing, the role is evolving towards a function closer to software architecture than to traditional programming. AI speeds up execution, but the developer provides the judgement, contextual knowledge, big-picture view and the ability to validate the final solution. This not only reduces development times but also allows human effort to be focused on what is truly important and critical. The result is twofold: greater speed and, at the same time, an improvement in the quality of the deliverable, thanks to a greater focus on design, validation and integration.
The importance of secure development in AI projects and business applications
If there is one area where this evolution is particularly relevant, it is that of secure development. In artificial intelligence projects and business applications, automation only delivers real value when it is integrated with secure design practices throughout the entire software lifecycle.
And this becomes even more important when AI is involved in generating, reviewing or correcting code, because validation must take place on two levels: not only must it be verified that the solution works and responds correctly to the stated need, but also that it is secure, maintainable, compliant with the defined architecture and aligned with compliance requirements. In other words, AI can speed up development, but it also places greater demands on the validation of the result.
Furthermore, the adoption of AI introduces a new risk surface that did not exist in traditional development. It is no longer sufficient to review only the AI-generated or AI-assisted code; one must also ensure how the AI itself is used within the process. OWASP has formalised this reality in its Top 10 for LLM Applications 2025, where it identifies specific risks resulting from the use of AI. In other words, as well as using AI to improve security, it is also necessary to govern and secure it as a technology with its own risks.
Therefore, when it comes to secure development, AI must be understood as a capability multiplier, not as a substitute for expert judgement. It can automate reviews, detect early warning signs and speed up rectification, but final validation still requires architectural context, business knowledge and human oversight. In fact, OWASP emphasises that manual review remains essential for detecting business logic issues, complex workflows and vulnerabilities that automated tools do not always capture, and Microsoft expressly warns that AI-generated fixes must be reviewed and tested before being integrated.
How AI improves vulnerability detection and strengthens software cybersecurity
AI can bring significant value to automated code review. When properly integrated into the development process, it enables the analysis of changes, the identification of software errors, the detection of potential vulnerabilities, the identification of insecure configurations, the review of dependencies and the proposal of fixes before these issues reach more advanced stages or even production. This type of early review helps to anticipate faults, reduce the number of changes and improve the quality of the final product. Current review tools, such as GitHub, already combine static analysis with AI capabilities to flag issues in pull requests, suggest fixes and speed up resolution, albeit always under the developer’s supervision.
In a landscape where applications are built and deployed at an ever-increasing pace, having support mechanisms that enable the anticipation of faults or weaknesses represents a significant advantage.
In other words, it can become a valuable aid in moving from a reactive approach to security towards a more proactive one. That said, this evolution requires clear governance frameworks, validation criteria and seamless integration with the organisation’s existing security tools and processes.
Key developer skills in the age of artificial intelligence
The incorporation of artificial intelligence into software development is profoundly redefining the skills a developer needs. It will remain essential to master fundamentals such as object-oriented design, concurrency, APIs, databases, service integration and performance. However, on top of this technical foundation, new key competencies are emerging: the ability to work with intelligent assistants, the judgement to assess the quality of generated solutions, an understanding of the risks associated with models, an awareness of privacy and data protection, and applied knowledge of cybersecurity and compliance.
Added to all this is an increasingly important requirement: the ability to adapt quickly to a rapidly changing paradigm. The way software is developed using AI is constantly evolving, with new tools, new capabilities and new ways of interacting with the technology. Therefore, beyond learning a specific tool, developers need to cultivate a flexible mindset that is open to change and geared towards continuous learning.
Furthermore, AI is introducing a new level of abstraction into the way software is built. In a sense, the developer is beginning to resemble a workshop master rather than someone who manually executes every detail. Just as it is said that Leonardo da Vinci would explain to his assistants what they should paint in a picture and reserve his own intervention for the truly decisive elements, today’s developer can give AI high-level instructions on what to build, how to structure it or which problem to solve, and focus their work on the important, complex or distinctive details.
In this context, one of the most valuable skills will no longer be simply knowing how to programme, but knowing how to supervise, direct and refine. The developer must understand when an AI-generated solution genuinely speeds up the work and when, conversely, it may introduce technical debt, logical errors, vulnerabilities or an implementation that is poorly aligned with the intended architecture.
Therefore, the developer of the future will need to combine technical depth with critical thinking, the ability to review code, architectural vision and a much more cross-functional understanding of the software lifecycle.
AI innovation with security and compliance: Telefónica’s approach
In a company such as Telefónica, innovation in artificial intelligence must always go hand in hand with robust frameworks for security, compliance and accountability. The adoption of these capabilities cannot be considered solely in terms of efficiency or the acceleration of development, but must also be based on clear criteria for governance, traceability, information protection and risk management.
In this regard, one of the key principles is to always work with secure and isolated tools and environments, so that data remains within the corporate sphere and can be managed with the appropriate guarantees of protection, confidentiality and compliance.
Therefore, the real difference lies not merely in adopting artificial intelligence, but in doing so responsibly and on a solid technological foundation. This means using it where it truly adds value, integrating it with mature engineering practices, defining clear controls and maintaining human oversight in critical processes.
Key future trends: generative AI, development and cybersecurity
In the coming years, we will see an even greater convergence between development, artificial intelligence and cybersecurity. One of the clearest trends will be the evolution towards models that are increasingly efficient, lighter and easier to integrate, enabling them to be incorporated into a growing number of development cycle workflows.
Furthermore, a significant part of this evolution will involve the growing use of on-premises models, run within the corporate environment itself or on infrastructure controlled by the organisation.
At the same time, these systems will become increasingly autonomous, evolving towards agent-based architectures capable of coordinating complex tasks. In parallel, we will see a growing specialisation in models specifically geared towards IT security.
All of this means that the future of development will not involve replacing the developer, but rather raising their level of abstraction and responsibility even further, with a growing focus on oversight, validation and governance.
The distinguishing factor: responsible innovation with a real impact on customers and society
In this context of technological transformation, the distinguishing value lies not merely in adopting artificial intelligence, but in doing so with a strategic, sustainable vision centred on the real impact for customers and society. Telefónica approaches this evolution from a clear standpoint: combining innovation, security, quality of service and corporate responsibility.
This approach translates into a commitment to technological leadership in Europe, the development of advanced services and the construction of infrastructure that enables citizens and businesses to access technology securely, efficiently and reliably. Telefónica has taken on the undeniable challenge of becoming the best gateway for citizens to access digital technologies
From this perspective, artificial intelligence, software development and cybersecurity are not isolated elements, but key pillars for building a more competitive, innovative and reliable digital ecosystem, where service quality is ultimately measured by customer trust.
