Artificial intelligence is increasingly recognized as a relevant resource for the energy sector, yet cross-sectoral evidence consistently places the industry among those with the lowest deployment maturity. A large share of firms remains in pilot or exploratory stages, with limited ability to scale solutions beyond initial proof of concept. The gap between declared investment and operational implementation suggests that the main constraint is not technological but organizational and informational: data quality, governance structures, human capital, and sourcing strategies determine whether AI generates scalable and capturable value or remains confined to isolated initiatives. This thesis examines the conditions under which AI becomes strategically relevant in the energy sector. A structured literature and sector review maps the state of the art in AI adoption, maturity, and applications, identifying four recurrent gaps related to scalability, value measurement, business model (BM) reconfiguration, and governance operationalization. Starting from these gaps, a strategic analysis investigates the exploitation and exploration balance, the organizational enablers of scalability, and the make, buy, or ally sourcing decisions. An empirical investigation then tests whether the patterns discussed in the strategic analysis are observable in practice, applying a content analysis to public documents issued by ten European firms, five from the energy sector and five from financial services, over the 2022 to 2024 triennium, using a five-dimensional coding scheme capturing Value Framing, Concreteness, Governance, Sourcing, and Domain. The results show that the energy sector follows a positive adoption trajectory but lags behind financial services on all principal indicators. Governance exhibits the sharpest cross-sectoral divergence, consistent with the proposition that it functions as a structural enabler of scalability rather than a compliance requirement. Sourcing profiles diverge from expectations, with financial services showing stronger internal development orientation and the energy sector relying more on partnerships. Both sectors display directional movement consistent with a commoditization logic in which firms progressively shift from exploitation toward exploration as AI capabilities mature. The findings are, however, based on strategic communication and do not capture operational deployment, which constitutes the main limitation of the analysis.
L'intelligenza artificiale viene sempre più riconosciuta come una risorsa rilevante per il settore energetico, eppure le evidenze comparative collocano l'industria tra quelle con la più bassa maturità di implementazione. Una quota consistente di imprese opera ancora in fase sperimentale, con una limitata capacità di portare a scala le soluzioni oltre i primi casi d'uso. Il divario tra investimenti dichiarati e implementazione operativa suggerisce che il vincolo principale non sia tecnologico ma organizzativo e informativo: qualità dei dati, strutture di governance, capitale umano e strategie di approvvigionamento tecnologico determinano se l'intelligenza artificiale genera valore scalabile e appropriabile. Questa tesi esamina le condizioni che rendono l'intelligenza artificiale strategicamente rilevante nel settore energetico. Una rassegna strutturata della letteratura e del settore identifica quattro lacune ricorrenti relative a scalabilità, misurazione del valore, riconfigurazione dei modelli di business e operazionalizzazione dei principi di governance. A partire da queste lacune, è stata svolta un'analisi strategica che tratta l'ambidestria organizzativa, le condizioni abilitanti per la scalabilità e le strategie di sourcing. Successivamente, un’indagine empirica verifica se i pattern discussi nell’analisi strategica siano riscontrabili nella pratica. A tal fine, è stata condotta un'analisi del contenuto sui documenti pubblici di dieci aziende europee, di cinque del settore energetico e cinque dei servizi finanziari, nel triennio 2022-2024, utilizzando uno schema di codifica a cinque dimensioni: orientamento strategico del valore, concretezza, governance, sourcing e dominio applicativo. I risultati mostrano che il settore energetico segue una traiettoria di adozione positiva ma rimane in ritardo rispetto ai servizi finanziari su tutti i principali indicatori. La dimensione della governance presenta la divergenza più marcata, coerentemente con la proposizione che essa funzioni come condizione abilitante e non come mero requisito di conformità. I profili di approvvigionamento divergono dalle attese, con i servizi finanziari orientati verso lo sviluppo interno e il settore energia più dipendente da collaborazioni esterne. Entrambi i settori si spostano progressivamente dallo sfruttamento verso l'esplorazione man mano che la maturità tecnologica aumenta. Tuttavia, i risultati si basano sulla comunicazione strategica e non catturano l'implementazione operativa, il che costituisce il principale limite dell'analisi.
Digital transformation: the strategic impact of Artificial Intelligence in the energy sector
Sesenna, Alessandro
2024/2025
Abstract
Artificial intelligence is increasingly recognized as a relevant resource for the energy sector, yet cross-sectoral evidence consistently places the industry among those with the lowest deployment maturity. A large share of firms remains in pilot or exploratory stages, with limited ability to scale solutions beyond initial proof of concept. The gap between declared investment and operational implementation suggests that the main constraint is not technological but organizational and informational: data quality, governance structures, human capital, and sourcing strategies determine whether AI generates scalable and capturable value or remains confined to isolated initiatives. This thesis examines the conditions under which AI becomes strategically relevant in the energy sector. A structured literature and sector review maps the state of the art in AI adoption, maturity, and applications, identifying four recurrent gaps related to scalability, value measurement, business model (BM) reconfiguration, and governance operationalization. Starting from these gaps, a strategic analysis investigates the exploitation and exploration balance, the organizational enablers of scalability, and the make, buy, or ally sourcing decisions. An empirical investigation then tests whether the patterns discussed in the strategic analysis are observable in practice, applying a content analysis to public documents issued by ten European firms, five from the energy sector and five from financial services, over the 2022 to 2024 triennium, using a five-dimensional coding scheme capturing Value Framing, Concreteness, Governance, Sourcing, and Domain. The results show that the energy sector follows a positive adoption trajectory but lags behind financial services on all principal indicators. Governance exhibits the sharpest cross-sectoral divergence, consistent with the proposition that it functions as a structural enabler of scalability rather than a compliance requirement. Sourcing profiles diverge from expectations, with financial services showing stronger internal development orientation and the energy sector relying more on partnerships. Both sectors display directional movement consistent with a commoditization logic in which firms progressively shift from exploitation toward exploration as AI capabilities mature. The findings are, however, based on strategic communication and do not capture operational deployment, which constitutes the main limitation of the analysis.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_03_Sesenna_thesis .pdf
solo utenti autorizzati a partire dal 22/02/2027
Descrizione: Thesis
Dimensione
1.21 MB
Formato
Adobe PDF
|
1.21 MB | Adobe PDF | Visualizza/Apri |
|
2026_03_Sesenna_Executive_Summary.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: Executive summary
Dimensione
323.33 kB
Formato
Adobe PDF
|
323.33 kB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/252347