our education on media literacy is lagging behind; especially when we consider how fast the media landscape is changing (AI generated content is a lot more systematic at emotional rhetoric than we think and can do it at an absolutely unprecedented scale).

An (AI-generated) example of how this could look: Let's say two outlets report on immigration policy. Both are factually accurate. Both cite the same statistics. But:
Outlet A: "New immigration framework announced, affecting approximately 2 million people annually"
Outlet B: "Devastating policy change threatens to upend lives of 2 million families"

Both statements contain the same core information. But the emotional loading is completely different. “Announced” versus “threatens”. “Affecting” versus “upend lives”. “People” versus “families”.

So in these simplified examples, they’re pre-framing mine and your emotional response. They're choosing loaded language, alarmist phrasing, and moral framing that activates specific emotional triggers before our rational brain even gets a chance to process the information.

With this in mind, our updated emotional media literacy needs to operate on two layers:

The Information Layer: Is this factually accurate? Are the sources credible/any vested interest? This is what some schools already teach, and it does matter.

The Emotional Layer: How is this information making me feel? What emotional response is being activated? Am I feeling angry, afraid, outraged, hopeful? And critically: was I meant to feel this way?

Essentially, I’m trying to make that second layer visible with my project, Ethoscore (ethoscore.org). It’s a machine-learning model and emotional media-literacy curriculum trained on 125k+ articles that gives you a score, a metric, that shows the user when language is being used to activate emotional responses rather than convey information neutrally. While the use of emotional rhetoric might be intuitive at times, other times it can be quite subtle.