Bank of England: AI poses new risks to financial stability

Emerging risks from the rise of artificial intelligence (AI) have been flagged by a number of respondents to the Bank of England’s bi-annual survey on systemic risks.

This report presents the results of the 2023 H2 survey, which was conducted between 14 August and 8 September. 56 firms participated in the 2023 H2 survey, representing a 65% response rate.

Survey respondents remain confident in the stability of the UK financial system and reported a similar level of confidence to 2023 H1. Indeed, the perceived probability of a high-impact event affecting the UK financial system in both the short term and medium term is lower than judged in the previous survey.

Cyber-attack and geopolitical risks remain the most frequently cited risks among participants. The number of participants citing risks associated with a UK economic downturn has continued to increase sharply. The number of survey respondents citing inflation risk has slightly increased after having decreased in the previous survey.

The risk of cyber- attack, geopolitical risk and inflation risk are still considered the most challenging for firms to manage by a significant margin.

The Bank of England noted that its financial stability objective is to protect and enhance the stability of the financial system of the United Kingdom. The Systemic Risk Survey contributes to this objective by quantifying and tracking, on a biannual basis, market participants’ views of risks to, and their confidence in, the stability of the UK financial system.

The survey is generally completed by executives responsible for firms’ risk management or treasury functions. The results presented are based on responses to the survey and do not necessarily reflect the Bank of England’s views on risks to the UK financial system. Participants include UK banks and building societies, large foreign banks, asset managers, hedge funds, insurers, pension funds, large non-financial companies and central counterparties. Summary statistics are calculated by giving equal weight to each survey response.