I’ve been receiving many, MANY, questions over the last year around what is and isn’t ethical when deciding whether and how to use AI. Two questions I’ve started asking myself are these:

  • Would I be comfortable being on the receiving end of this?
  • Would I want this for my loved ones, like my niece and nephew?

Call this the “Recipient Test.”

This simple question isn’t just about deciding whether to use AI or not; it also prompts us to consider how we use it responsibly and ethically.

Take the example of recommendation letters. Are you tempted to use AI to draft them? Imagine seeing a letter written for you and discovering it was generated by an algorithm that knows nothing about you. Worse, picture your child receiving such a letter at a pivotal moment. This makes us question the wholesale outsourcing of the task. However, the “Recipient Test” can also illuminate a more thoughtful approach, one shared by my brilliant friend and colleague Tonia A. Dousay. Imagine someone who writes a few of these letters a week. How might theυ use AI responsibly? They might provide the job description, the candidate’s key qualifications, and their own personal insights to a tool AI, then dedicate time to revising and personalizing the draft. In this way, the “Recipient Test” can lead us to use AI as an assistant rather than a replacement.

Consider grading and providing feedback – this is a treacherous terrain. Sure, it saves time, but I still remember truly personalized feedback that shaped my learning. Would an AI provide that? I also recall receiving generic feedback – not from an AI – that offered little value. Perhaps the “Recipient Test” here encourages us to think about how AI can augment, rather than fully automate, feedback, allowing educators to focus on providing individualized guidance. I’m suspect of the “saving us time” argument, but that’s a whole different post.

When designing learning activities, you might apply this test by asking: Would using AI to generate a learning activity create a memorable learning moment I’d like to participate in? Would it inspire curiosity in my niece? The “Recipient Test” might compel you to look beyond mere efficiency.

The use of AI in education – like the use of so many other technologies over the decades – is about thoughtful, measured, and critical integration. The test might lead us to conclude that AI is or isn’t appropriate in a particular case, but it also pushes us to define the parameters of its use to ensure a positive and ethical outcome for the recipient. By consistently asking this question, we are forced to pause, reflect on the impact, and make informed choices.

Aside: This kind of thoughtful consideration might be more challenging in cases where AI is embedded within tech/edtech products, where some of our agency is limited.