The ACM Code of Ethics and Data Science
The General Moral Imperatives of the ACM Code of Ethics serve as a basic guideline for computing professionals to remain honest and confidential in their work. We’ve seen how the advancements in the use of data to make decisions has had a positive impact on technology, however, we must follow some rules to ensure the safety of people who use our products as well as the environment and society we contribute to.
The first moral imperative listed in the code of ethics is the contribution to society and human well-being. Though it may seem obvious the role of computing professionals is to advance and improve technology for all people, the first moral imperative encourages professionals to make sure we leave a safe and habitable environment. Though an essential ethical issue, one area often forgotten by those in technology is the proper disposal of outdated technologies, in particular, batteries. Lithium-ion batteries power most of our mobile devices and we’ve seen how dangerous they can be when defective, such as the Samsung Galaxy S7 explosions. We need to ensure standards in mass battery disposal, which often happens when company’s upgrade systems and recycle outdated products.
The ACM Code of Ethics also warns professionals to “avoid harm to others”. Undoubtedly the longest portion of the moral imperatives cautions professionals from causing harm, intentionally and unintentionally. The Code of Ethics also outlines means to mitigating unintentionally harm. This ethical issue is exemplified in white-hat versus black-hat hacking. Hacking by definition violates the code of ethics, and more often than not is used as a means to cause harm, by either stealing credit card or other personal information. However, there are companies who hire white-hat hackers to expose deficiencies in their security systems, thus avoiding data breaches.
The moral imperative of honesty and trustworthiness are applicable in Data Science specifically, in how our data is represented. We have the power to design biased experiments and display data in misleading ways, and it remains our ethical duty to maintain integrity, transparency, and honesty in our work. A great example of this is the Rocky Mountain High-Intensity Drug Trafficking Area’s, “Legalization of Marijuana” report. The report claims there was a 169% increase in cannabis exposure for children ages 0-5, however, the number of exposures only grew by 22 children. The report also claims there has been a 67% increase in roadway fatalities involving cannabis, despite flawed THC tests and the unreliability of their results. The highlights of the report are highly misleading and as professionals, we’re called on to eliminate bias and dishonesty in our system and experimental design.
The Code of Ethics calls on professionals to operate fairly and to take action against discrimination. As technology advances, we need to look further than our systems directly discriminating against people. As Airbnb and Uber learned in recent years, though their technology doesn’t inherently work against anyone based on race, sex, religion, age, etc., it allows their users to do so very easily. Though their systems weren’t based on discrimination, a responsible course of action is to work against these unfair practices and the burden is on the company to reduce the ways discrimination can creep onto their platforms. As an issue of social justice, both companies have implemented anti-discrimination policies.
One of the greatest issues facing the technological world is the issue of who owns what or digital property rights. The Code of Ethics outlines that computing professionals should avoid violating copyright, patent, and license agreements. It also cautions us only to duplicate software with express authorization. It’s not entirely difficult for skilled professionals to scrape a website and steal parts if not all of their code. Many professionals resort to duplicating software because of how expensive professional-level software can be. Students often do this for school projects but should be respectful of the work and effort put in to create software and pay developers fairly for their work. As an Ethical issue, stealing another’s work wouldn’t be accepted in any industry. More so, with intellectual property laws protecting many works, we could face serious legal repercussions for violating copyrights or unauthorized software duplication.
In addition to honoring property rights and patents, the code of ethics explicitly states to give credit for intellectual property. As Data Scientists we all rely on data often compiled or formatted by other people. It’s rare that a professional is in a position like myself, where we manage the collection as well as the analysis. In such cases, correct attributions need to be made in order to stay within ethical practices.
A major issue facing our increasingly digital world is the issue of privacy which to me, is closely related to consent. Professionals need to protect user privacy at all costs because even if the data breached seems “harmless”, it could have dire implications for users. As we navigate the new industries advancements in technology helped shape, we have to manage storing and protecting the vast amounts of user data collected in order to use data-driven decisions to contribute to society. A great example of bad data security is the bevy of cyber attacks Sony has faced. From the PlayStation Network hack of 77 million personal details including credit card numbers and the 2014 Sony Pictures personnel hack that exposed millions of internal emails and private employee data, Sony has proven to have low-security standards. The Japanese company has failed to protect their user and employee data from “unauthorized access or accidental disclosure to inappropriate individuals” as specified in the Code of Ethics. This is a major privacy issue as a company who is entrusted by millions of customers to provide beyond a modicum of respect for their data.
The last moral imperative listed in the Code of Ethics drives home the point that professionals should be honest and honor confidentiality promises. More and more modern technologies use data users didn’t expressly consent to companies using. For example, demographic data such as age, household income, and gender as well as psychographics are used to target ads to users on sites like Facebook and Twitter, based on their general browser history. While some of this data is compiled by classifying users based on past searches by Facebook and Twitter themselves, most often it’s bought from a larger corporation that users don’t consent to giving their information to. This is a huge issue of privacy and we as budding data scientists specifically need to respect the anonymization of personal data when conducting analyses.
Overall the ACM Code of Ethics’ first section lays out eight moral imperatives for computing professionals to obey. These imperatives cover the scope of Ethics, Privacy and Social Justice.