‘Artificial Intelligence and Systemic Risk’ discussed with Jón Daníelsson (LSE)
In the last online seminar by the Florence School of Banking and Finance, we discussed the risks and opportunities of artificial intelligence with Jón Daníelsson, Director of the Systemic Risk Centre at the London School of Economics.
Artificial intelligence (AI) is rapidly changing how the financial system is operated, taking over core functions because of cost savings and operational efficiencies. In this online seminar, Professor Daníelsson discussed the positive and negative effects that these changes may introduce in the sector, and how they will affect the role and tasks of risk managers and microprudential authorities.
In particular, Professor Daníelsson highlighted the differences between how AI can be applied in micro and macro prudential supervision. In microprudential suppervision and risk management, despite the technology to perform most of the functions would be already present, there are still relevant obstacles posed by cultural, political and legal considerations. In the case of macroprudential supervision, there are still technological limits to AI, which are reflected particularly in its inability to identify “unknown unknowns” (i.e. events that have not been imagined).
Finally, Professor Daníelsson discussed some of the aspects that can help reducing the risks brought by AI, advocating in particular for a ‘kill switch’ or, in other words, the possibility for human intervention to override the decisions taken by an AI. Asked to express their opinion on this feature, 96% of the participants agreed with the speaker.