Generative AI is radically transforming the digital economy. After shaking up the way we use technology, it is now taking on the very heart of business productivity – software development. ‘Vibe coding’, a method for producing code through interactive conversation with AI, is emerging as a symbol of this transformation. Being faster, more accessible and sometimes less technical, it promises to accelerate digital creation. However, behind this promise of greater efficiency lies a veritable skills shake-up, which companies are still struggling to fully grasp. 

 

Rapid Adoption 
The magnitude of this phenomenon is already evident. According to a report ‘The State of AI in Software Engineering 2025′, over 63% of organisations state that, to deliver code, they use AI assistants more frequently than not. Development and engineering teams use between eight and 10 distinct AI tools. Adoption is widespread and rapid, yet sometimes sporadic. AI-assisted development is no longer a niche experiment; it is becoming a new operating standard. 

 

In this context, vibe coding is lowering the barrier to entry. It enables less experienced users, and even business teams, to generate features with just a few instructions. Software creation is becoming more accessible. For senior management, the benefits are clear: reduced time-to-market, increased product testing, and enhanced in-house innovation. Code is becoming an accessible strategic lever. 

 

However, the report also sheds light on a paradox. Although 51% of coding workflows are automated, the rate drops to 43% for build pipeline creation and execution. The upstream is accelerating, whilst the downstream is struggling to keep up. The result is tangible: 45% of all deployments involving AI-generated code leads to problems, and 72% of organisations have already experienced at least one production incident relating to the use of AI-generated code.

 

Acceleration can lead to fragility. For companies, the risk goes beyond the technical sphere alone. It affects cybersecurity, regulatory compliance, and reputation. It also has an impact on costs with some 70% of organisations worried about potentially spiralling cloud expenditure linked to AI-generated code that ultimately proves inefficient. In other words, the promise of greater efficiency may conceal an invisible increase in risks and costs. 

 

The shake-up is primarily one of skills. Developers are no longer simply code writers. They are becoming system architects, responsible for ensuring the robustness of pipelines and for supervising generative algorithms. Savoir-faire is now geared towards designing secure environments, automating testing, and standardising deployments. In fact, 57% of organisations consider it a priority to standardise build and deployment processes to provide a framework for AI usage. Value now lies in mastering the system rather than in production itself. 

 

Need Safety Measures  
This shift is generating internal tension. The widespread use of code can lead to a greater reliance on the most skilled developers, who will need to correct the code, make it secure, and optimise it. Nearly two-thirds of respondents believe that vibe coding could be ‘a disaster waiting to happen’ if the appropriate safety measures are not put in place. Companies cannot simply just adopt the tool; they will also need to restructure their organisations. 

 

Maturity is emerging as a differentiating factor. Organisations capable of automating the entire software lifecycle, from writing to deployment, succeed in combining speed and resilience. Others risk finding themselves in a danger zone that is fast but fragile. 

 

Vibe coding does not spell the end for developers; rather, it marks the end of a model based exclusively on manual development. In today’s digital economy, competitive advantage will no longer depend on the ability to write code faster, but rather on the ability to orchestrate AI intelligently, to make processes secure, and to reconcile innovation with risk management. 

 

The revolution is not so much technological as organisational. It requires companies to invest in training, governance, and standardisation. Generative AI is accelerating digital transformation, but this, in turn, calls for an equally profound transformation of skills. 

 

(The author is chief innovation and digital officer, Excelia and an expert in digital transformation and AI) 

Source link