About the research paper (Tucci, with G. Lanzolla and S. Santone): 

"One of the sectors that AI is expected to transform, or even disrupt, more radically is the financial services sector. For instance, Hawksworth et al. (2018) indicate that the jobs most vulnerable in the earliest stages of AI automation (the “algorithmic wave”) are in financial services. In this paper, we set out to shed some light on the mechanisms through which AI can transform/disrupt value creation in the financial sector, and on why extracting value is proving harder than expected.

Thus, first, we develop a conceptual model of AI value creation through the lens of the business model construct. Then, we undertake a study of how artificial intelligence has been associated with transformation/disruption in the financial services sector over the 2013-2019 time period. To deliver on this goal, we use a topic modeling approach on two sources of data: a “corpus” of articles published in the global business press (in the English language) concerning AI in financial services; and a corpus of articles published on web pages concerning AI in financial institutions located in the City of London, one of the major financial centers in the world. The topic modeling on the global business press corpus of data reveals general (global) narratives about AI in financial services. The second study on the Internet corpus of data reveals the narratives related to financial institutions located in the City of London.

Finally, we compare and contrast our model of AI’s value creation with the narratives emerging from our topic models. Our analyses reveal that there is a significant divide between the strategic and organizational tradeoffs that should be made to use AI for value creation and the narratives around AI in the media.

We conclude that perhaps one of the reasons for AI still being a long way from delivering on its potential is the chasm between the tough choices necessary to leverage AI for value creation and these somewhat generic—or even overly simplistic—narratives. We offer suggestions on how to close this divide by commenting on AI adoption timing and market entry timing".