Financial Stability Board voices concerns over AI “black boxes”

Maria Nikolova

Financial regulators, institutions and consumers are facing the problem of interpretability of AI decisions in areas like trading and investment, FSB warns.

The Financial Stability Board has voiced its concerns about possible risks associated with the increased adoption of artificial intelligence (AI) and machine learning in the financial services sector. In a report entitled “Artificial intelligence and machine learning in financial services. Market developments and financial stability implications”, the FSB notes the numerous advantages of such novel technologies but also highlights some risks that may stem from their use.

One of the biggest risks is that the use of AI and machine learning may create the so-called “black boxes” in decision-making that could result in complicated issues, especially during tail events. In particular, it may be difficult for humans to understand how decisions, such as those for trading and investment, have been formulated.

Given this problem with interpretability, if AI and machine learning based decisions cause losses to financial intermediaries across the financial system, there may be a lack of clarity around responsibility.

The use of complex algorithms could result in a lack of transparency to consumers when it comes to credit scores assigned by AI programs. When using machine learning to assign credit scores make credit decisions, it is generally more difficult to provide consumers, auditors, and supervisors with an explanation of a credit score and resulting credit decision if challenged. Additionally, some argue that the use of new alternative data sources, such as online behaviour or non-traditional financial information, could introduce bias into the credit decision.

With regard to the use of such novel technologies in portfolio management, where AI and machine learning tools are used to identify new signals on price movements and to make more effective use of available data and market research than with current models, one issue is that useful trading signals derived from AI and machine learning strategies may follow a decay function over time, as data are more widely used and hence become less valuable for gaining an edge over other investors.

Another risk stems from a situation where multiple market participants come to use similar AI and machine learning programs in areas such as credit scoring or financial market activities. If machine learning-based traders outperform others, this could in the future result in many more traders adopting similar machine learning strategies. As with any herding behavior in the market, this has the potential to increase financial shocks. Moreover, advanced optimisation techniques and predictable patterns in the behavior of automated trading strategies could be used by insiders or by cybercriminals to manipulate market prices.

High frequency trading (HFT) applications of AI and machine learning could be new sources of vulnerabilities. If a similar investment strategy based on AI and machine learning is widely used in HFT, it might amplify market volatility through large sales or purchases executed almost simultaneously.

Finally, network effects and scalability of new technologies may in the future give rise to third-party dependencies. This could in turn result in the emergence of new systemically important players that could fall outside the regulatory perimeter.

  • Read this next


    Volt secures EMI license, expands payment solutions in UK

    Volt has successfully obtained an Electronic Money Institution (EMI) license from the UK’s Financial Conduct Authority (FCA).

    Retail FX

    ASIC bankrupts finfluencer Tyson Scholz over stock tips

    The Australian Securities and Investments Commission (ASIC) has effectively bankrupted Tyson Robert Scholz, the figure behind “Black Wolf Pit.” The action marks a significant crackdown on so-called ‘finfluencers’ and individuals providing unlicensed financial services.

    Digital Assets

    Green Bitcoin Presale Raises $1M as Bitcoin Approaches its ATH

    The eco-friendly crypto project Green Bitcoin has seen its limited-time presale phase cross $1 million in funding. With an innovative gamified staking model and energy-efficient foundation, Green Bitcoin offers token holders a way to stake their tokens and generate yield.


    Introducing QuickNode Streams: Elevating Blockchain Data Management

    Discover QuickNode’s Latest Innovation: Streamlining Blockchain Data Streaming for Enhanced Efficiency and Accessibility. Explore the Future of Blockchain Technology with Streams.

    Industry News

    John Oliver rips into MetaTrader over role in ‘Pig Butchering’ scams

    “If your friend told you to download an app, and you saw it in the app store with good reviews, you might assume everything on it was legitimate. In before, you saw MetaTrader’s logo which looks like three men in suits jerking each other off under a table – an appropriate metaphor for cryptocurrency if I have ever seen one,” Oliver quipped.

    Digital Assets

    Coinbase supports Nethermind and Erigon to ease Geth dependency

    Coinbase plans to support additional execution clients as America’s largest crypto platform aims to improve the Ethereum blockchain’s resilience and mitigate the risks associated with the network’s heavy reliance on a single client.


    How AI Transforms Trading: Current Trends and Perspectives

    In 2023, we observed a boom of news about Artificial Intelligence (AI) in every field, whether finance, tech or medicine. In 2024 and later, AI will take an even more significant place.

    Industry News, Uncategorized

    FCA wants to tackle lack of competition in wholesale data market

    “Complex licensing practices by MDVs and trade data providers who deliver their data through MDVs increase costs for data users. Many Market Data Vendor (MDV) users have to hold licences both from the data generator (such as a trading venue) and from the MDV through which they access data. We have seen an increasing proliferation of licences for similar data types and different use cases. Complexity also drives additional costs for data users, such as operating a compliance team.”

    Digital Assets

    SEC objects to Terraform’s $166 million legal retainer

    The U.S. Securities and Exchange Commission (SEC) has lodged objections against Terraform Labs for a $166 million retainer payment to its legal representatives ahead of its trial.