Generative AI Adoption in an Energy Company: Exploring Challenges and Use Cases

Organisations are examining how generative AI can support their operational work and decision-making processes. This study investigates how employees in a energy company understand AI adoption and identify areas where AI and LLMs-based agentic workflows could assist daily activities. Data was collected in four weeks through sixteen semi-structured interviews across nine departments, supported by internal documents and researcher observations. The analysis identified areas where employees positioned AI as useful, including reporting work, forecasting, data handling, maintenance-related tasks, and anomaly detection. Participants also described how GenAI and LLM-based tools could be introduced through incremental steps that align with existing workflows. The study provides an overview view of AI adoption in the energy sector and offers a structured basis for identifying entry points for practical implementation and comparative research across industries.

💡 Research Summary

This paper presents an empirical case study of generative AI (GenAI) adoption within a mid‑size Nordic energy company. Over a four‑week period in early 2025, the authors conducted sixteen semi‑structured interviews with senior staff and unit leads across nine functional areas (customer operations, core infrastructure, information systems & data integration, finance & reporting, market operations, asset & risk planning, workforce development, strategic projects, and business development). The interview protocol captured participants’ roles, daily challenges, existing tools, and expectations or concerns regarding AI. Complementary data were gathered from internal documents and on‑site observations, providing triangulation for the qualitative analysis.

Through iterative coding, the researchers identified 41 distinct AI‑related use cases. These were grouped into five primary work domains: (1) reporting automation, (2) forecasting (demand, price, market trends), (3) data handling and integration, (4) maintenance‑related tasks (predictive maintenance, diagnostic assistance), and (5) anomaly detection. Each use case was evaluated on three dimensions: business importance (high, medium, low), implementation difficulty (easy, moderate, hard), and expected organizational value (cost reduction, productivity gain, risk mitigation).

The analysis revealed that reporting and forecasting present the highest business impact while also being relatively easy to implement because they can leverage existing Business Intelligence (BI) platforms and APIs. Maintenance support, especially a Retrieval‑Augmented Generation (RAG)‑based chatbot that surfaces equipment manuals and past failure records, scored moderate on difficulty but offered substantial risk‑reduction benefits. Anomaly detection, though valuable, requires real‑time streaming data pipelines and raises data‑privacy/compliance concerns, making it the most challenging to pilot initially.

Based on the scoring, the authors prioritized two pilot projects: (1) an automated reporting and forecasting tool that integrates with the company’s ERP/BI stack, and (2) a maintenance‑assistant chatbot built on RAG techniques. Both pilots include a staged rollout plan—initial prototype, user training, performance monitoring, and iterative refinement—spanning roughly six months.

In addition to the technical findings, the study maps organizational barriers to AI adoption: data silos, legacy system incompatibility, stringent regulatory constraints, and limited AI literacy among staff. To mitigate these, the authors recommend establishing a cross‑functional AI governance team, adopting a “pilot‑then‑scale” approach anchored in measurable ROI, and launching targeted internal up‑skilling programs.

The paper contributes to the literature by moving beyond individual‑level productivity studies and offering a holistic, organization‑wide perspective on GenAI integration in a knowledge‑intensive, highly regulated sector. It also provides a replicable methodology for other firms seeking to identify, prioritize, and pilot AI use cases. Limitations include the single‑company scope, modest interview sample size, and reliance on self‑reported data, which may affect external validity. Future work is suggested to evaluate the pilots’ actual impact, conduct longitudinal ROI analyses, and compare adoption pathways across different industries.