Can Administrators Ensure the Ethical Use of AI in K12 Education? – EdTech Magazine: Focus on K-12

As with many technologies used in K12 learning environments, school leaders must guarantee that artificial intelligence is safe. In addition to the legal requirements placed on districts, there are ethical issues schools must consider before introducing new tech powered by AI and machine learning (ML).

To do this, there must first be an understanding of what AI looks like in education. All of us have very different ideas for AI and algorithmic tools and machine learning and what these things are, sayssava saheli singh, an eQuality-Scotiabank postdoctoral fellow in AI and surveillance at the University of Ottawa.

The concepts of AI and ML are nebulous, meaning that everyone understands them a bit differently. The internet, and companies selling AI-powered products, can sway peoples understanding of this technology.

For school leaders to properly protect students using this tech, they must understand what AI and ML mean in the context of the products theyre adopting. IT administrators should also understand how the AI is using data and how the tools affect different student populations.

Click the bannerto learn more about trending tech when you register as an Insider.

The first consideration leaders must make is whether they understand the tools in which theyre investing. IT departments should know what the tools are doing behind the scenes to produce the results they see.

Theres definitely a lack of understanding about what these systems are, how theyre implemented, who theyre for and what theyre used for, singh says.

In understanding how AI works, IT leaders must remember that these systems rely on data.

AI only works if you grab data. It only works if you grab data from everywhere you can find it, and the more data the better, saysValerie Steeves, a professor in the department of criminology at the University of Ottawa and principal investigator of theeQualityproject.

Because AI and ML tools rely on data, admins must ensure theyre building in student data protections to use this technology ethically. AI thats rolling out now always comes with a price tag, and the price tag is your students data, Steeves says.

KEEP READING:Data analytics shows the impact of educational technology.

Another ethical consideration is algorithmic bias in AI and ML technologies. When purchasing these solutions, its important to remember that these programs are operating off data sets that frequently contain bias.

Theres a lot of bias involved in applying some of these tools, singh says. A lot of these tools are made by specific people and with specific populations in mind. At a basic level, theres racial, gender, sexual orientation differences theres a lot of different kinds of people and a lot of these technologies either leave them out or include them in ways that are really harmful.

While the creators of these tools typically arent intending to cause harm, the biases built into the data sets can discriminate against certain student populations.

It creates a system where biases can play out in in ways that are rampant, and it becomes ever more difficult to pull them back because the bias and the discriminatory use of the data is built into the algorithm, Steeves says.

One way to account for this bias is to acknowledge it. Teaching students to interact with AI should include lessons about how it was created and where these biases may appear.

The necessary ethical considerations shouldnt dissuade school leaders from using AI in education entirely. Teaching students to interact with this technology will set them up for success in higher education and future careers. There are also safe ways to use AI.

A lot of the context in which algorithms and AI and machine learning are useful is when youre looking at a large corpus of data and trying to make sense of it, singh says. Using algorithms and AI to answer a specific question can maybe give you a clue as to what the larger context might be. In that educational context, I think its a useful tool.

When students are inputting data, instead of being the subjects from whom data is extracted, AI can be extremely beneficial.

AI is most useful when no data is collected from the kids, and the kids are not embedded into some kind of surveillance system, Steeves says. AI is just helping them facilitate their learning and their modeling of the world around them.

DIVE DEEPER:Learn how to use artificial intelligence in K12 education.

View original post here:

Can Administrators Ensure the Ethical Use of AI in K12 Education? - EdTech Magazine: Focus on K-12