Optimising AI-driven cyber defence for European energy infrastructure through economic risk modelling and cyber diplomacy
Year: 2026
Type: article
Abstract: European electricity systems increasingly depend on interconnected digital control infrastructure while artificial intelligence (AI) is being introduced into cyber-threat detection. This study examines the economic and energy implications of AI-assisted cyber defence in the European energy sector and the relevance of cross-border cyber diplomacy for operational resilience. It integrates publicly documented indicators from ENISA, the International Energy Agency, Eurostat, and IBM with a transparent scenario-based decision model for a hypothetical critical-energy operator. ENISA analysed 4,875 cybersecurity incidents during July 2024–June 2025, of which 1.7% were classified in the energy sector. IEA estimates indicate European data-centre electricity use of 68 TWh in 2024 and 114 TWh under its 2030 Base Case. The decision model assumes a 15% annual probability of a significant incident, an illustrative EUR 2 million impact per incident, a EUR 60,000 annual fixed AI-defence cost, and a 10 kW incremental IT load. Under these assumptions, the required reduction in expected incident loss to break even is 27.6%. A 40% reduction yields an annual net benefit of approximately EUR 37,136, whereas a 25% reduction remains below break-even. All economic results are conditional simulations rather than measured field performance. The findings support risk-based procurement, energy metering, human oversight, and trusted cross-border incident-information exchange as complementary instruments of cyber resilience and green economic optimisation.
DOI: 10.68210/jgeor.a100
Author: Zulfiyya Mustafayeva
Institution: Azerbaijan State Oil and Industry University
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7233603940
FWCI:
Cites:
Cited by:
Related to:
