{"id":2415,"date":"2019-10-02T11:37:32","date_gmt":"2019-10-02T11:37:32","guid":{"rendered":"https:\/\/sanlaw.legal\/?p=2415"},"modified":"2019-10-02T11:37:38","modified_gmt":"2019-10-02T11:37:38","slug":"to-err-is-human-but-who-is-responsible-for-machine-made-decisions","status":"publish","type":"post","link":"https:\/\/sanlaw.legal\/he\/to-err-is-human-but-who-is-responsible-for-machine-made-decisions\/","title":{"rendered":"To ERR is Human\u2026 BUT Who is responsible for Machine-Made decisions?"},"content":{"rendered":"\n

The second decade of the 21st<\/sup>\ncentury will probably be marked by future historians as the dawn of the\nArtificial Intelligence (AI) era. While we are yet to be hunted by legions of\nkiller robots guided by an AI resolved to correct god\u2019s mistake of saving Noah\nfrom the flood, we no longer use machines just to substitute or enhance human\nphysical labor, but also as a substitute for human discretion in the\ndecision-making process.<\/p>\n\n\n\n

With much media attention\nfocusing on autonomous vehicles making moral decisions in choosing between\npassengers and bystanders\u2019 safety, it is easy to forget that whereas autonomous\nvehicles or self-guided drones are still under development, other AI machines are\nalready making decisions which affect our everyday lives. <\/p>\n\n\n\n

In machines we trust?<\/strong><\/p>\n\n\n\n

Whether they employ the\nlatest AI technologies like artificial neural networks (ANNs), or a simple\nold-fashioned if-then flowchart algorithm, we use machines (which we call \u201ccomputers\u201d)\ndaily to make decisions for us and replace our human discretion. Lost on your\nway or just seeking to avoid traffic? A navigation software can decide on the\nbest route for you. Too busy to sort through your e-mails? A spam filter can\ndecide which ones to keep and which ones to discard. Trying to figure out what\nto watch? An App can analyze your past choices to decide which movie you may\nlike. <\/p>\n\n\n\n

Who is responsible when you\nare led to a traffic jam and late for a meeting with your boss? Who is to blame\nwhen that important message from a potential client is overlooked because it\nwas filed in the junk folder, or for an evening wasted on a boring or\ndistasteful movie? Who is liable for the result of a wrong decision when there\nis no human is involved in the decision-making process?<\/p>\n\n\n\n

The answer is easy; the\nperson deciding to rely on a machine decision for his or her convenience can be\nmade to agree to assume the risks of a wrong decision, in exchange for such\nconvenience. The answer gets more complicated when the subjects of machine-made\ndecisions have no choice or are unware of the use of machines in the\ndecision-making process and especially when the decisions might have a greater\neffect on their life.<\/p>\n\n\n\n

Need a loan? You can most\nlikely get one online without any human involvement. A computer will calculate\nyour credit history and asset value to evaluate your risk factor and based on\nthis factor, determine the amount you can get and the interest rate you would\nbe offered. The same is true for deciding what will be the insurance premium;\nwhich applicant should be admitted in a coveted school; or even when deciding\non the amount of bail money a suspect would need to post to avoid detention. In\nsuch cases, the operator of the decision-making machine makes use of the\nmachine to make many decisions affecting many individuals, to whom the\nliability for erroneous decision cannot be transferred. <\/p>\n\n\n\n

The ContentID™ algorithm used\nby YouTube™ to identify copyright infringement has been the focus of numerous\nlawsuits for falsely identifying original or public domain creations as\ninfringing. The Correctional Offender Management Profiling for Alternative\nSanctions software \u201cCOMPAS\u201d used by US Courts to asses flight risk and set bail\nis being criticized and challenged for being bias against certain minorities.\nBiometric Facial Recognition systems are slammed for being inaccurate and too\nvulnerable and easy to deceit. False, discriminatory, unfair, inaccurate or\notherwise wrong decisions, exposing the ones who rely thereon or act upon them\nto claims, are not exclusive to humans and in fact are much more common in\ndecision-making machines. <\/p>\n\n\n\n

Machines make decisions, not responsibility.<\/strong><\/p>\n\n\n\n

A citation attributed to\nAmerican scientist Paul Ehrlich \u201cTo err is human, but it takes a computer to\nreally foul up things\u201d<\/em> sums things up well. The advantages of having a\nmachine making multitude decisions faster and cheaper than any human, can\neasily become a disadvantage when the machine gets it wrong. In such a case,\ntraditional defenses applicable to human decision makers, such as the mistake\nbeing an isolated incident, a deviation from the organization\u2019s policy, bias or\nmalice by the said personal, exceeding his or her authority or acting on their\nown, etc. cannot be applied to a machine without attributing the responsibility\ndirectly to the operator thereof. Thus, the multitudinous of decisions on one\nhand and the direct responsibility of the operator to each of them on the other\nhand, increases the operator exposure to undesirable results such as class\nactions, negative media, damage to reputation, inquires by consumer protection\nauthorities, increased regulation, etc.  <\/p>\n\n\n\n

Obviously, the operator of\nthe decision-making machine can seek indemnification from the provider or\ndeveloper of the machine. However, in most cases the operator and the developer\nare the same entity, or the operator itself is involved in adjusting or training\nthe machine. <\/p>\n\n\n\n

In such cases there are measures that can be taken to\nminimize or mitigate the exposure:<\/p>\n\n\n\n

Transparency<\/strong>: As\nwith any human made decision, disclosing the decision-making process and the\ncriteria for making thereof, makes the decision less arbitrary, more\npredictable, and reduces the frustration of the person or people affected by\nit. Even in cases where full transparency is not possible for reasons such as\nproprietary decision-making technologies, a partial transparency, in the form\nof reasoning the decision or explaining what term or criteria were met or not\nmet by the subject of the decision, is preferable. Transparency can demonstrate\nthat the decision, although made by a machine, was not arbitrary, biased,\ndiscriminatory or otherwise unfair. Indeed, some AI technologies, such as ANN,\npose challenges in implementing such transparency.  <\/p>\n\n\n\n

Option to Appeal<\/strong>:\nOffering subjects of machine-made decisions the option to appeal decisions they\nbelieve to be erroneous to a human referee, even if such an appeal involves\ncosts and bureaucratic procedure, can serve to shift some responsibility for\nthe error in the machine-made decision process, from the operator of the\nmachine to the subjects. This is because by exercising their discretion in\ndeciding whether or not to appeal, and why, the subjects of the decision are no\nlonger totally passive and thus share, at least in part, the responsibility of\nthe final outcome of the decision. Of course, the appeal process must be\nreasonably available and the human referee authorized and capable of reversing\nor amending the decision in cases where the appeal is justified. <\/p>\n\n\n\n

Alternatives<\/strong>: When subjects of decisions are given a choice whether the decision in their cases will be made by human or by machine, they can also be requested to assume the risks of erroneous decisions by the machine, in exchange for the benefits of receiving the decision faster, for free, etc. In conclusion, with emerging AI technologies making machine-made decisions more and more common, relying on such decisions may increase exposure to liability for wrong decisions when they occur. Operators of machine making decisions and the ones relying on those decisions should be made aware of these potential exposures and take measures to minimize or mitigate them. <\/p>\n","protected":false},"excerpt":{"rendered":"

The second decade of the 21st century will probably be marked by future historians as the dawn of the Artificial Intelligence (AI) era. While we are yet to be hunted by legions of killer robots guided by an AI resolved to correct god\u2019s mistake of saving Noah from the flood, we no longer use machines […]<\/p>\n","protected":false},"author":1,"featured_media":2416,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"footnotes":""},"categories":[1],"tags":[],"class_list":["post-2415","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-uncategorized"],"acf":[],"aioseo_notices":[],"_links":{"self":[{"href":"https:\/\/sanlaw.legal\/he\/wp-json\/wp\/v2\/posts\/2415","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/sanlaw.legal\/he\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/sanlaw.legal\/he\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/sanlaw.legal\/he\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/sanlaw.legal\/he\/wp-json\/wp\/v2\/comments?post=2415"}],"version-history":[{"count":5,"href":"https:\/\/sanlaw.legal\/he\/wp-json\/wp\/v2\/posts\/2415\/revisions"}],"predecessor-version":[{"id":2421,"href":"https:\/\/sanlaw.legal\/he\/wp-json\/wp\/v2\/posts\/2415\/revisions\/2421"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/sanlaw.legal\/he\/wp-json\/wp\/v2\/media\/2416"}],"wp:attachment":[{"href":"https:\/\/sanlaw.legal\/he\/wp-json\/wp\/v2\/media?parent=2415"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/sanlaw.legal\/he\/wp-json\/wp\/v2\/categories?post=2415"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/sanlaw.legal\/he\/wp-json\/wp\/v2\/tags?post=2415"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}