Meticulous Transparency—An Evaluation Process for an Agile AI Regulatory Scheme

Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10868)


Artificial intelligence (AI) poses both great potential and risk, as a rapidly developing and generally applicable technology. To ensure the ethical development and responsible use of AI, we outline a new ethical evaluation framework for usage by future regulators: Meticulous Transparency (MT). MT allows regulators to keep pace with technological progress by evaluating AI applications for their capabilities and the intentionality of developers, rather than evaluating conformity to static regulations or ethical codes of the underlying technologies themselves. MT shifts the focus of ethical evaluation from the technology itself to instead why it is being built, and potential consequences. MT assessment is reminiscent of a Research Ethics Board submission in medical research, with required explanation depending on the potential impact of the AI system. We propose the use of MT to transform AI-specific ethical quandaries into more familiar ethical questions, which society must then address.


Assessment framework AI ethics Intentionality Meticulous transparency 



We are grateful to Dr. Wendell Wallach (Yale University) and to Dr. Jason Behrmann (aifred health), whose comments improved this article.


  1. Asilomar AI Principles. Accessed 24 Oct 2017
  2. The Bioss AI Protocol. Accessed 16 Jan 2018
  3. Bjornsdottir, R.T., Rule, N.O.: The visibility of social class from facial cues. J. Pers. Soc. Psychol. 113(4), 530–546 (2017)CrossRefGoogle Scholar
  4. Chen, C., Seff, A., Kornhauser, A., Xiao, J.: DeepDriving: learning affordance and direct perception in autonomous driving. In: IEEE International Conference on Computer Vision, pp. 2722–2730. IEEE Computer Society, Washington, DC (2015)Google Scholar
  5. Hardt, M., Price, E., Srebro, N.: Equality of Opportunity in Supervised Learning. arXiv:1610.02413 (2016)
  6. Hughes, G.: Montreal AI pioneer warns against unethical uses of new tech. CBC News (2017)Google Scholar
  7. Artificial Intelligence–The Public Policy Opportunity. Accessed 16 Oct 2017
  8. The IEEE Global Initiative for Ethical Consideration in Artificial Intelligence and Autonomous Systems. Accessed 20 Jan 2016
  9. Information Technology Industry Council AI Policy Principles. Accessed 20 Sept 2017
  10. Open Letter on Autonomous Weapons. Accessed 29 Oct 2017
  11. Jarcho, J.M., Berkman, E.T., Lieberman, M.D.: The neural basis of rationalization: cognitive dissonance reduction during decision-making. Soc. Cognit. Affect. Neurosci. 6(4), 460–467 (2011)CrossRefGoogle Scholar
  12. Kavukcuoglu, K., Sermanet, P., Boureau, Y.-L., Gregor, K., Mathieu, M., Cun, Y.L.: Learning convolutional feature hierarchies for visual recognition. In: Advances in Neural Information Processing Systems, vol. 1, pp. 1090–1098 (2010)Google Scholar
  13. Larson, J., Mattu, S., Kirchner, L., Angwin, J.: How We Analyzed the COMPAS Recidivism Algorithm. ProPublica (2016)Google Scholar
  14. Luo, W., Li, Y., Urtasun, R., Zemel, R.: Understanding the effective receptive field in deep convolutional neural networks. arXiv:1701.04128 (2017)
  15. Nguyen A, Yosinski J, Clune J.: Deep neural networks are easily fooled: high confidence predictions for unrecognizable images. arXiv:1412.1897(v4) (2015)
  16. Pan, H., Jiang, H.: A deep learning based fast image saliency detection algorithm. arXiv:1602.00577 (2016)
  17. Russell, S., Dewey, D., Tegmark, M.: Research priorities for robust and beneficial artificial intelligence. AI Magaz. 36, 105–114 (2015)Google Scholar
  18. Tiku, N.: Welcome to the Next Phase of the Facebook Backlash. Wired (2017)Google Scholar
  19. Wallach, W., Allen, C.: Moral Machines: Teaching Robots Right from Wrong. Oxford University Press, New York (2010)Google Scholar
  20. Wallach, W.: Toward a ban on lethal autonomous weapons: surmounting the obstacles. Commun. ACM 60(5), 28–34 (2016)CrossRefGoogle Scholar
  21. World Medical Association: World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA 310(20), 2191–2194 (2013)CrossRefGoogle Scholar
  22. Zeng, D., Chen, H., Lusch, R., Li, S.H.: Social media analytics and intelligence. IEEE Intell. Syst. 25(6), 13–16 (2010)CrossRefGoogle Scholar
  23. Zevenbergen, B., Mittelstadt, B., Véliz, C., Detweiler, C., Cath, C., Savulescu, J., Whittaker, M.: Philosophy meets internet engineering: ethics in networked systems research. In: GTC Workshop Outcomes, pp. 1–37. Oxford University, Oxford (2015)Google Scholar
  24. Zoph, B., Yuret, D., May, J., Knight, K.: Transfer learning for low-resource neural machine translation. arXiv:1604.02201 (2016)
  25. Montavon, G., Braun, M.L., Müller, K.-R.: Kernel analysis of deep networks. J. Mach. Learn. 12, 2563–2581 (2011)MathSciNetzbMATHGoogle Scholar
  26. Zeiler M. D., Fergus, R.: Visualizing and understanding convolutional networks. arXiv:1311.2901(v3) (2013)

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  1. 1.McGill UniversityMontrealCanada
  2. 2.aifred healthMontrealCanada
  3. 3.Douglas Mental Health University InstituteMontrealCanada
  4. 4.Montreal Neurological InstituteMontrealCanada

Personalised recommendations