Building Good Digital Citizens: Uncovering Bias With Machine Learning and Artificial Intelligence

Share on facebook
Share on twitter
Share on linkedin

Artificial intelligence (AI) was typically only found in dystopian science fiction novels, but that is no longer the case. Everything from disease recognition and prevention to security at large events, AI is making its way into our everyday lives and its presence will only grow as technology advances. 

Due to its rapid growth, it’s important for any educator to identify how students will be impacted by AI and how we can prepare them for the future. Although there are many facets of AI that will soon, if not already, be incorporated into our daily lives, this article will focus on one important aspect that is facial and image recognition.

What is AI Image Recognition?

AI image recognition uses a subset of AI called machine learning that has the ability to identify and categorize specific objects, including people, animals, and various landmarks. Utilizing a wealth of data, computers can use a camera and a machine learning algorithm  based on neural networks and pre-trained data to identify these objects.

In order for a computer to make sense of images they must be broken down into pixels, with each pixel having a value (0 = black / 255 = white). Below is a photo showing how an image is broken down into pixels and how each pixel has a value.

What Issues Does AI Image Recognition Have?

AI isn’t perfect and is only as good as the data it has at its disposal. If a machine learning image recognition tool struggles to recognize the difference between two different images, it’s likely because it does not have the data, or images, necessary to make an accurate distinction between the two. 

For instance, as infants we learn to see patterns in colors and shapes so that we can recognize them later. We start to get good at predicting based on what we have already seen. For AI to be good at predicting, the software needs to have seen lots of images and examples.

When an AI image recognition tool does not have lots of data, or does not have a variety of data, predictability issues arise and the tool is not as effective as one with more information. This can lead to inaccurate predictions and a bias that could have significant consequences to society as a whole as this tool becomes more widespread.

How Can We Educate Students About AI Bias?

As AI image recognition tools become more utilized for a variety of purposes, it’s important our students recognize the common issues with these tools and help prevent the negative aspects that can present themselves. Even providing students with a basic understanding of what AI is, how it works, and what limitations it has can go a long way in helping our students make informed decisions. 

As a part of Computer Science Education Week and in conjunction with Hour of Code, BSD Education is offering a free project that allows students to explore an AI image recognition tool and teaches them how AI sees images, makes predictions, and how we can make sure that those predictions are accurate and free of bias. While utilizing this tool, students will also learn valuable 21st century coding skills, including the basics of HTML, CSS, and JavaScript.

Access AI Webcam Image Classifier at

We hope you and your students will find this project informative and helpful in understanding how AI works and can be used for good in the future. BSD Education feels strongly that this type of learning, along with other digital skills, is vital to the education of every student. If you would like to learn more about how you can partner with BSD in the classroom so your students are future ready, please contact us here.

Want to Learn More?

Free BSD Webinar

Building Good Digital Citizens: Uncovering Bias With Machine Learning

Join BSD Education VP of Education Mark Barnett to see how AI sees images, makes predictions, and how we can make sure they are accurate and free of bias. We’ll also explore similarities between AI and how our own minds work by studying the effects of bias and categorization of data.

about Beth
Beth is the Marketing Manager at BSD Education.
She is focused primarily on full-stack marketing strategy, data analytics, and email marketing at BSD. Beth has over eight years of experience and has worked with several industries globally, and operates a full-stack marketing consultancy based in Hong Kong.

other posts in


Join our

Get news, teaching ideas and resources from the BSD Education community every 2 weeks.


BSD Education (Build Something Different) partners with educators to bring technology education into classrooms through our programs of learning, online learning platform and professional development training.

Join our Free Insights Newsletter!
Get free teaching resources and tips into your inbox