[Digital Today reporter Yoonseo Lee] As companies speed up adoption of artificial intelligence (AI), many developers appear to use tools to demonstrate how much AI they use.

TechRadar reported on Oct. 7 that consulting firm Adaptavist confirmed the trend in a survey of 1,000 software developers in Britain, the United States, Germany, the Netherlands and Spain.

According to the survey, 50 percent of respondents said they use AI because they must prove they can use it. Another 47 percent said they feel their work is judged by how much AI they use rather than the quality of the output. Companies are putting large budgets into AI adoption, but on the ground, whether AI is used is becoming a more important metric than actual efficiency.

The pressure is also reflected in performance evaluation systems. Some 58 percent said AI usage has become a performance measurement criterion or has been added to key performance indicators (KPIs). As organisations push AI use without fully considering existing work methods and development processes, it means developers are feeling more anxious about what role they should play within the organisation.

Job insecurity also emerged. Some 54 percent said they feel anxious about long-term career prospects, and 37 percent said they worry they could lose their jobs if they do not use AI. It suggests AI is being seen not as a tool to boost productivity but as a factor affecting job stability and evaluations.

Matt Saunders (매트 손더스), head of DevOps at Adaptavist Group, said a beneficial development experience is important not only for individual well-being but also for organisational growth and innovation. He pointed out that AI adoption goals, usage metrics and a reward system that values a culture of being seen using AI rather than using it well should not harm the experience.

He then warned such a structure could lead to talent departures. He argued that organisations should look at results rather than fixating on adoption rates in the process of embracing AI.

The task for companies is shifting from AI expansion itself to how to manage the results of its use. Saunders said that to retain engineering talent, companies should stop measuring AI adoption rates and focus on outcomes, and give developers time, autonomy and support to use AI in ways that actually help improve their work. As AI becomes a new evaluation item for development organisations, whether companies can move away from usage-centred management will be a key point to watch.