Abstract
Using the question of how to treat algorithmic evidence under the Confrontation Clause as an entry point, I argue that the use of AI in ethically salient situations presents a risk. It may cause us to avoid confronting our own responsibility. This matters because facing up to what we do, including what we delegate, can help us grow and thrive. Bearing responsibility can help us nurture vital capacities, including forms of empathy, honesty, and dignity. In the language of ethics, these are eudaimonist virtues—traits and capacities that can help us live well and fully. We should thus find ways of ensuring that we bear the ethical weight of what we do. I call such methods conscience catchers. These are practices that, in addition to the pragmatic work they do, also serve to affix a sense of felt responsibility. In the context of the criminal trial, the Confrontation Clause is an example. But there are many ethically salient circumstances beyond the trial in which it will increasingly make sense to rely on the help of AI. Some, such as investment decisions and marketing, although perhaps less dramatic than criminal trials, involve many more people. We should thus look for conscience catching devices in those contexts also. Having identified and argued for the significance of the risk that reliance on AI may lead us to avoid confronting responsibility, I wish I could describe precisely how to guard against that risk. I cannot. The problem has not been sufficiently recognized or studied. I can, with examples from areas I have studied and inspiration from literature, describe some sorts of methods that might serve a conscience catching function. But beyond that I call upon lawyers, policy makers, and social scientists to help us find ways to catch our conscience in the many situations in which we can and ought to carry the weight of what do, including what we delegate to our algorithms.
Recommended Citation
Sherman J. Clark,
Confronting Algorithms: Conscience Catching in the Criminal Trial and Beyond,
57
U. Mich. J. L. Reform
787
(2024).
Available at:
https://repository.law.umich.edu/mjlr/vol57/iss4/6
Included in
Artificial Intelligence and Robotics Commons, Criminal Law Commons, Criminal Procedure Commons, Legal Ethics and Professional Responsibility Commons