Three lines of defense against risks from AI

Schuett, Jonas

doi:10.1007/s00146-023-01811-0

Cited by 5 publications

(2 citation statements)

References 58 publications

(60 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is worth noting that some companies already have a function which is responsible for assessing the adequacy and effectiveness of their risk management practices: internal audit [146,147]. If the company already has an internal audit function, the additional value of an ethics board with similar responsibilities would be limited.…”

Section: B Reviewing Risk Management Practicesmentioning

confidence: 99%

See 1 more Smart Citation

How to design an AI ethics board

Schuett,

Reuel,

Carlier

2024

AI Ethics

Self Cite

View full text Add to dashboard Cite

The development and deployment of artificial intelligence (AI) systems poses significant risks to society. To reduce these risks to an acceptable level, AI companies need an effective risk management process and sound risk governance. In this paper, we explore a particular way in which AI companies can improve their risk governance: by setting up an AI ethics board. We identify five key design choices: (1) What responsibilities should the board have? (2) What should its legal structure be? (3) Who should sit on the board? (4) How should it make decisions? (5) And what resources does it need? We break each of these questions down into more specific sub-questions, list options, and discuss how different design choices affect the board’s ability to reduce societal risks from AI. Several failures have shown that designing an AI ethics board can be challenging. This paper provides a toolbox that can help AI companies to overcome these challenges.

show abstract

Section: B Reviewing Risk Management Practicesmentioning

confidence: 99%

“…In addition to that, they need sound risk governance [15,92]. For example, they may establish a board risk committee, appoint a chief risk officer (CRO), and set up an internal audit function [146,147]. In this paper, we explore yet another way in which AI companies can improve their risk governance: by setting up an AI ethics board.…”

Section: Introductionmentioning

confidence: 99%

How to design an AI ethics board

Schuett,

Reuel,

Carlier

2024

AI Ethics

Self Cite

View full text Add to dashboard Cite

show abstract

Risk and artificial general intelligence

Faroldi

2024

AI & Soc

View full text Add to dashboard Cite

Artificial General Intelligence (AGI) is said to pose many risks, be they catastrophic, existential and otherwise. This paper discusses whether the notion of risk can apply to AGI, both descriptively and in the current regulatory framework. The paper argues that current definitions of risk are ill-suited to capture supposed AGI existential risks, and that the risk-based framework of the EU AI Act is inadequate to deal with truly general, agential systems.

show abstract

Crossing the lines a human approach to improving the effectiveness of the three lines model in practice

Fenelon,

van Doorn,

Scholten

2024

JFRC

View full text Add to dashboard Cite

Purpose Financial services firms have a significant societal responsibility to prevent issues. The three lines model helps them do that though faces challenges in its effectiveness. This paper aims to offer a behavioural perspective on these challenges and practical solutions to help improve the model and herewith better prevent issues. Design/methodology/approach The authors detail key behavioural pitfalls and underlying psychological mechanisms that hinder the effectiveness of the model. The authors illustrate these with examples from the corporate practice, alluding to the behavioural patterns and drivers identified in the academic and consultancy work. The authors conclude with offering practical solutions how to enhance the effectiveness of the model. Findings The authors discuss common ineffective intergroup behaviours between the controllers (here: internal audit) and the controlled (here: the audited business or 2nd line functions): the controllers responding to issues with increased scrutiny; the controlled dismissing the feedback and challenging the issues raised; and the controlled and the controller competing for power. The root causes of these ineffective intergroup behaviours include: psychological defence mechanisms, social categorisation and collective beliefs about intrusiveness. The offered solutions range from actions the controllers can take, actions the controlled can take and actions both can take to improve the effectiveness of the model in practice. Originality/value The authors argue that the behavioural perspective on the effectiveness of the model is a blind spot and largely omitted from organisations’ agendas. This paper adds this behavioural perspective to help organisations improve the effectiveness of the model.

show abstract

Three lines of defense against risks from AI

Cited by 5 publications

References 58 publications

How to design an AI ethics board

How to design an AI ethics board

Risk and artificial general intelligence

Crossing the lines a human approach to improving the effectiveness of the three lines model in practice

Contact Info

Product

Resources

About