Subscribe
Article Complete
AI Features

Why Data Science Companies Need Chief Ethics Officers Now More Than Ever

Published

Why Data Science Companies Need Chief Ethics Officers Now More Than Ever

Key takeaways

  • Emerging technologies present significant ethical challenges that companies are struggling to address.
  • Machines learn from data, which can lead to biased outcomes if the input data is flawed.
  • Instances of AI displaying racial bias highlight the necessity for ethical oversight in technology.
  • The role of Chief Ethics Officers is becoming increasingly important in addressing data ethics.
  • Denial among tech professionals regarding data ethics has contributed to the prevalence of biased algorithms.

Takeaways drafted with AI from the article text.

Despite all the advancements and benefits that emerging technologies have brought recently, there are significant challenges that companies across the world are trying to cope — ethical challenges.

When Machines Become Biased

No matter how much we emphasise the advancement of machines, at the end of the day, we cannot deny the fact that machines learn with the help of data that is being fed. And the actions it would take will be also be based on the same data. In 2016, in a beauty contest, a robot called Beauty.AI was given a chance to judge the contest, thinking an artificial intelligence would be unbiased and do justice. However, to the surprise, the robot turned out to be racist — out of 44 winners, it had selected only one winner with dark skin. This is not the only case when an AI has become racist, even companies like Microsoft are not exempt from it. In 2016, Microsoft unveiled an AI-based Twitter bot called Tay as an experiment in "conversational understanding", but it took Twitter less than 24 hours to turn into a racist. As soon as Microsoft launched the chatbot, Twitter users bombarded the bot with all sorts of misogynistic, racist tweets and it seemed the bot took things literally and learnt everything and started replying back with the same flavour tweets.

The Need For Chief Ethics Officer

When it comes to data ethics, most techies seem to be in denial. This has resulted in biased algorithms that sometimes end up impacting the organisation as well as individuals. However, that shouldn’t be a case.  As the ethical challenges in the data science domain continue to grow across the world, data ethics is becoming a crucial field of study. And to make data ethics much more serious, organisations (that mostly work with data science) across the world have started to onboard chief ethics officer, and they are emphasizing the fact that the need is real. Now, Salesforce has hired an ethics chief, Paula Goldman, in January 2019. To be specific, Goldman carries the title of chief ethical and humane use officer at Salesforce.  Another example of a company is Google. The search giant might not have a dedicated officer to take care of data ethics, but it has a board that is focused entirely on ethics and AI. Who is a Data Science Chief Ethics Officer In general, a chief ethics officer is someone who is responsible for helping in understanding the difference between what is a right and a wrong outcome. Simply put, they develop and distribute codes of ethics and professional conduct. They are also responsible for developing training programs for employees, and monitoring and auditing compliance with government regulations. However, in terms of data science, the role is much more focused on the domain. A chief ethics officer makes sure that there is no ill-suited use of data. The officer also keeps things very clear among professionals about the type of data that needs to be fed to the machine. The Role Of An Ethics Officer It is no surprise that data science is a vast domain, and a professional cannot always keep in mind about all the aspects of a project. And this why a data science chief ethics officer is allotted to take care of that.  A chief ethics officer makes sure that every time there is a project ready to go live, it takes into consideration the two most important question:  Who will be benefited and who will be affected by this project? Is this project going to do injustice to any community? Once you get the answers to the questions and if they align well to your organisation’s vision, then it is good to go. For example, when you feed historical data to a machine, you are basically letting the machine to repeat the past. So, in that case, you should not only feed data to the machine but also analyse the result and see whether it is biased, whether the data is right, what other kinds of data should also be fed etc.

Outlook

Ethics is one of the less talked about things in data science, and as a result, it has impacted a lot (negatively) in the past few years. There are instances of machines being biased because of data, and if companies are still not considering taking a look back at their data ethics, then it is high time that starts doing it. And also consider onboarding a Chief Ethics Officer who would exclusively take care of the data ethics aspect.

Published

Corrections: AIM corrects errors of fact and notes the change on the story. Spotted one? Tell the editors. Read our editorial standards.

Follow AIM
Preferred SourceGoogle NewsWhatsAppTelegram

More in this story

Data Science · 5 stories
  1. The War Between Python And Go Explained In 8 Points
  2. Top Indian Quora Writers To Follow For Data Science
  3. You are here: Why Data Science Companies Need Chief Ethics Officers Now More Than Ever
  4. Why India Is Such An Important Market For Data Science Product Vendors
  5. Unlocking 20% Growth: How India's Data Centre Surge is Reshaping Tech Services
About the author

Contributor

All stories
Belamy · Daily

The day’s AI news, before your first meeting.

Related