In this photograph taken on January 10, 2024, a traffic policeman directs traffic in Bhutan’s capital Thimphu. - AFP
THIMPHU: A recent survey by the Royal Civil Service Commission (RCSC) and the GovTech Agency has revealed a significant gap in Artificial Intelligence (AI) literacy and capability among civil servants in the Professional and Management Category.
The findings, based on 2,757 responses from 54 agencies, highlight an urgent need for targeted training and strategic workforce planning. This need is pressing even though 47 per cent of respondents believe AI can improve government efficiency.
The survey found that 27.1 per cent of civil servants use AI frequently, while 40.6 per cent use it occasionally.
The survey aimed to evaluate current AI awareness, skills and readiness among civil servants to identify skill gaps, assess training needs and develop learning programmes.
A major concern is the discrepancy between AI usage and formal training, since 67.7 per cent of respondents use generative AI tools like ChatGPT without having received any formal instruction.
The findings also revealed low awareness of AI ethics and responsible practices, alongside significant skill gaps in applying AI to specific professional domains.
Despite these skill deficits, interest in upskilling is high, with about 98 per cent of respondents expressing interest in practical AI training. However, the path to wider adoption faces several key barriers.
The report identified the main obstacles as a lack of structured training programmes (21 per cent), limited technical expertise (14 per cent), and inadequate infrastructure (14 per cent). A lack of in-house technical experts, coupled with budget constraints, was also identified as a major obstacle.
The report recommends developing a national AI infrastructure roadmap, securing funding or partnerships, and establishing departmental AI focal points supported by regular training and expert collaboration.
To address these deficits, the RCSC recommends a national strategy focussed on targeted training and policy integration. The primary solution involves designing a tiered AI training roadmap catering to all proficiency levels.
The beginner level would focus on basic AI literacy, the intermediate level on responsible AI use and public service applications, and the advanced level on hands-on AI project design, machine learning deployment, and leadership in AI strategy and governance.
The report states that AI training must be integrated into the civil service learning framework and made part of mandatory professional development.
It also states that training content must evolve to cover ethics, data privacy and prompt engineering, ensuring alignment with national laws and RCSC-GovTech Generative AI Guidelines. - Kuensel/ANN
