Writing with artificial intelligence? Ad-hoc-survey findings raise awareness for critical literacy at the International Literacy Day

Abstract

ABSTRACT The International Literacy Day 2023 was the first after the launch the text generating artificial intelligence ChatGPT. This was the reason for a Literacy Promptathon that allows users to learn how to handle text and image generation. The International Literacy Day media coverage for the first time touched on the question of AI generated text. One of the difficulties of generative AI is that it does not necessarily produce true information. However, findings show that critical literacy is most difficult for adults with low literacy performance. Research shows that disinformation, algorithmic biases and data collection are among the most relevant dangers of generative AI, while controlling these dangers requires precise reading and builds on high literacy skills. The research question was how users handle the risk of disinformation and whether this varies by experience. Data collection used an ad-hoc online survey with 202 respondents. Findings indicate an overall high awareness to avoid fake news, but not necessarily adequate fact-checking habits, especially among less experienced AI users. While limited to the respondents, the findings give reason to assume that there is a substantial need to improve competences in fact-checking. Adult and continuing education have a pioneering role in this regard.