Generative artificial intelligence (GenAI) is rapidly transforming the world of work, and the economics profession is no exception. For educators, this presents an urgent challenge: we need to understand how AI is reshaping what economists do in practice; what tasks are changing, what new capabilities are emerging and what skills are becoming essential. Only by doing so can we adapt our curricula and rethink how we assess students in an AI-driven world
This is precisely the aim of our research, which investigates how professional economists working in the corporate sector, the public sector and academia are currently using AI in their day-to-day work, and explores the implications for the employability skills of our future graduates.
We adopt a mixed-methods approach to capture both the breadth and depth of how economists use AI. We ran an exploratory survey of 114 professional economists working in various sectors in the spring of 2025, and complemented this with qualitative insights from an employer panel and focus groups held as part of a dedicated workshop.
Together, these elements allow us to identify both quantitative patterns and richer reflections on how AI is being integrated into economic work. This research was part of a wider research project on AI and assessment in economics, funded by the Stone Centre at UCL, which contributed to the Royal Economic Society (RES) report on rethinking economics assessments for a GenAI world.
GenAI is embedded across all layers of economic work
Our findings reveal that professional economists are deploying GenAI tools across a broad range of cognitive tasks. At one end, GenAI is used for low-level thinking tasks such as transcription and code debugging. At the other, it is increasingly involved in higher-order tasks like data analysis, problem-solving and creative exploration. This wide spectrum of use shows that GenAI is integrated into every layer of economics work.
Source: Nassehi et al (2025)
This diversity carries important implications for teaching and assessment. First, students must be supported to engage with GenAI at all levels, from basic automation to advanced problem-solving.
Second, assessment design should move beyond rewarding polished outputs towards valuing the full process of working with GenAI: how students define a problem, develop prompts, evaluate AI-generated content, iterate thoughtfully and justify their methodological choices.
‘Learning to learn’ with GenAI as the essential employability skill
The most important employability skill in the age of AI is not a single fixed capability, but the broader, layered meta-competence of ‘learning to learn’ with GenAI. This involves three interrelated elements:
- First, students need to develop fluency with the evolving capabilities of GenAI, such as coding, scripting and prompt engineering.
- Second, they must learn to think critically with GenAI: questioning outputs, recognising limitations and exercising evaluative judgement.
- Third, they should be able to apply GenAI tools in new and unfamiliar contexts, adapting them to diverse problems and environments.
These latter two elements are deeply connected to core research and problem-solving skills.
Source: Nassehi et al (2025)
Ethical and responsible use of GenAI as a new soft skill
As GenAI tools become more common in economic work, our findings point to the growing importance of soft skills that support ethical and responsible use. These include knowing when it is appropriate to use GenAI, understanding the importance of transparency, handling data with care and being aware of issues such as fairness and privacy.
Concerns raised by pulse survey participants and focus group contributors – ranging from over-reliance and deskilling to unequal access and concerns about the integrity of outputs – highlight the need to equip students with the habits and mindset to use GenAI thoughtfully in different professional settings.
Source: Nassehi et al (2025)
Taken together, this means that as economics educators we need to rethink fundamentally how we design teaching and assessment by focusing on three key priorities:
- First, we must find ways to integrate GenAI across all levels of teaching, including lectures, tutorials, assessments and learning activities. Crucially, we must ensure that students engage actively with it as part of their learning journey.
- Second, we need to assess students’ ability to work with AI by focusing on how they approach tasks – how they frame problems, interact with AI tools and evaluate outputs – rather than judging only the final result.
- Finally, it is essential that we explicitly embed and model the responsible and ethical use of GenAI within our teaching and assessment.
AI is here to stay, within economic work and in the wider world. It is high time that the profession starts building this new reality into its frameworks of teaching and assessment.
