AI and Social Media

baowenliu
1 Comment

Over the past couple years, the use of AI on social media has become increasingly widespread, particularly in media images, videos, and text. While this has enriched online information, it has also made it increasingly complex. People often struggle to distinguish between AI-generated content and reality, which can lead them to be misled by false information. An article by MediaSmarts notes: Deepfakes and other AI-generated misinformation are becoming impossible to avoid online—according to a recent report, two-thirds of Canadians were exposed to this kind of content multiple times in the past year（Survey of Online Harms in Canada，2025). AI-generated content has spread across the internet; we cannot avoid it, so we must learn to coexist with it.

One AI example I’d like to mention involves Air Canada’s chatbot. In 2022, Air Canada’s chatbot promised a discount that wasn’t available to passenger Jake Moffatt, who was assured that he could book a full-fare flight for his grandmother’s funeral and then apply for a bereavement fare after the fact.  According to a civil-resolutions tribunal decision last Wednesday, when Moffatt applied for the discount, the airline said the chatbot had been wrong – the request needed to be submitted before the flight – and it wouldn’t offer the discount （Yagoda, 2024). The final court ruling in this case was that Air Canada must reimburse Moffatt for the discounted amount, and that all information provided by AI on Air Canada’s official website shall be legally binding.

Image from godsavethepoints.com

This case highlights two issues. First, the information provided by AI makes it difficult for us to distinguish between fact and fiction; people seem to view AI-generated information as authoritative and place too much trust in it. This reminds me of how I, too, place a great deal of trust in ChatGPT when I ask it questions—because its tone is confident and its logic seems sound, I rarely question its accuracy. Second, the misinformation provided by AI can lead us to make wrong decisions and cause us financial or emotional losses. Therefore, this is not a mistake that can be ignored, but rather an issue that demands serious attention.

In light of these issues and given the difficulty of identifying AI, I believe we need to start by examining ourselves. I propose developing a media literacy course module focused on AI trustworthiness, encouraging students to reflect on why they trust AI and addressing the issue at its root. For example, do we unconditionally trust AI because of its confident, fluent language? We could then set a correct answer, have the AI provide the incorrect answer in fluent language, and have a human provide the correct answer in slightly halting language. By experiencing this contrast, students would learn that language style does not equate to authenticity.

References

Survey of Online Harms in Canada 2025. (2025, May 23). The Dais. https://dais.ca/reports/survey-of-online-harms-in-canada-2025/

‌Yagoda, M. (2024, February 23). Airline held liable for its chatbot giving passenger bad advice – what this means for travellers. Www.bbc.com. https://www.bbc.com/travel/article/20240222-air-canada-chatbot-misinformation-what-travellers-should-know

‌

One Comment

  1. Hi i enjoyed reading your blog. I like how you used the Air Canada chatbot example because it shows how errors in AI can actually cost people money and not just spread misinformation. Your own reflection on trusting AI, and how confident language can actually contribute to this, is really honest. I also think your idea for the activity on media literacy is really good because it challenges people to question the credibility of both wrong and right. This is because one is well spoken and the other isnt. It was really good to be reminded to question the information we receive from AI, not just consume it.

Comments are closed.