Abstract
AI systems that offer social services, such as healthcare services for patients, driving for travellers and war services for the military need to abide by ethical and professional principles and codes that apply for the services being offered. We propose to adopt Requirements Engineering (RE) techniques developed over decades for software systems in order to elicit and analyze ethical requirements to derive functional and quality requirements that together make the system-to-be compliant with ethical principles and codes. We illustrate our proposal by sketching the process of requirements elicitation and analysis for driverless cars.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
Notes
- 1.
But note, there are ethical requirements that are not legal, and legal ones that are not ethical.
- 2.
In [9], the focus is on use value as opposed to ethical value. However, we believe the analysis still holds, in particular, regarding the connection between value and risk.
- 3.
Notice that transparency w.r.t. the entities that compose an ecosystem regarding their capabilities, intentions, vulnerabilities, and goals strongly connects also to the notion of trust. In a nutshell, trust amounts to a set of relations connecting the beliefs of a (trustor) agent regarding the capabilities, vulnerabilities and intentions of a trustee insomuch as they can affect that agent’s goals [3]. From this we directly have that: (1) trustworthiness assessment can and should be grounded in the explicit assessment of these aspects; (2) trustworthiness is not an absolute property of a system, but one that depends on all these aspects. To put it bluntly, it is meaningless to speak of trustworthy systems in an unqualified manner.
References
Autonomy in weapon system, DoD directive. Technical report, Department of Defence (2012). https://tinyurl.com/vh2qhej
Artificial intelligence: Mankind’s last invention (2019). https://tinyurl.com/y9moo26c
Amaral, G., Sales, T.P., Guizzardi, G., Porello, D.: Towards a reference ontology of trust. In: Panetto, H., Debruyne, C., Hepp, M., Lewis, D., Ardagna, C.A., Meersman, R. (eds.) OTM 2019. LNCS, vol. 11877, pp. 3–21. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-33246-4_1
Arkin, R.: Lethal autonomous systems and the plight of the non-combatant. AISB Q. 137, 1–9 (2013)
High-Level Expert Group on Artificial Intelligence, E.C: Draft ethics guidelines for trustworthy AI, draft document (2018)
Etzioni, A., Etzioni, O.: Incorporating ethics into artificial intelligence. J. Ethics 21(4), 403–418 (2017). https://doi.org/10.1007/s10892-017-9252-2
O’Connell, M.E.: Banning autonomous killing. In: Evangelista, M., Shue, H. (eds.) The American Way of Bombing: How Legal and Ethical Norms Change. Cornell University Press, Ithaca (2013)
Otto, P.N., Antón, A.I.: Addressing legal requirements in requirements engineering. In: Proceedings of the 15th IEEE RE 2007, New Delhi, 15–19 October 2007, pp. 5–14 (2007)
Sales, T.P., Baião, F., Guizzardi, G., Almeida, J.P.A., Guarino, N., Mylopoulos, J.: The common ontology of value and risk. In: Trujillo, J.C., et al. (eds.) ER 2018. LNCS, vol. 11157, pp. 121–135. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-00847-5_11
Tuffley, D.: At last! the world’s first ethical guidelines for driverless cars. The Conversation, September 2017. https://tinyurl.com/u4gbskh
Acknowledgments
This research is supported by the Strategic Partnership Grant “Middleware Framework and Programming Infrastructure for IoT Services”.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Guizzardi, R., Amaral, G., Guizzardi, G., Mylopoulos, J. (2020). Ethical Requirements for AI Systems. In: Goutte, C., Zhu, X. (eds) Advances in Artificial Intelligence. Canadian AI 2020. Lecture Notes in Computer Science(), vol 12109. Springer, Cham. https://doi.org/10.1007/978-3-030-47358-7_24
Download citation
DOI: https://doi.org/10.1007/978-3-030-47358-7_24
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-47357-0
Online ISBN: 978-3-030-47358-7
eBook Packages: Computer ScienceComputer Science (R0)