AI Finds Influx of Vaping Posts on Social Media

A study led by USC researchers used artificial intelligence (AI) to scan the popular social media app TikTok and found a stark increase in vaping content from 2019 to 2022.

Despite regulations by the federal government, a settlement with JUUL and pledges by both social media companies and some e-cigarette companies not to promote vaping products to youth, AI found the prevalence of pod devices in promotional TikTok videos increased by 15 percent since 2019.

The instances of TikTok posts with images of e-juices increased by 33 percent from 2021 to 2022. The frequency of e-juice flavor names and vaping brand names increased by about 100 percent from 2019 to 2022.

The TikTok topic “#vaping” alone has been viewed 4.8 billion times since inception.

“Posts that promote tobacco are prohibited by platforms like TikTok and Meta, but studies show that only a small portion of that content gets removed,” said lead author Dr. Julia Vassey, a USC researcher and the lead author of the paper.

Vassey and Harvard Medical School Researcher Dr. Chris Kennedy used a form of AI known as computer vision to track the prevalence of various tobacco-related objects on social media. The results were recently published in the journal Nicotine & Tobacco Research.

Researchers started by building a computer vision model, which uses AI to identify specific features in visual data, such as photos or videos. Using a dataset of 6,999 images from Instagram, they trained the algorithm to recognize objects related to e-cigarette use.

Once the model was trained to recognize tobacco-related objects, Vassey and her team used it to analyze 14,072 TikTok videos posted from “micro-influencers.” These users have between 1,000 and 100,000 followers and get a high number of likes and comments on their posts.

The researchers have also publicly shared their model, code and training dataset for students and researchers to download and use on their own datasets.

TikTok, Meta and other platforms could use computer vision to identify tobacco-related imagery and remove it promptly, the researchers said, which could reduce youth exposure to vaping in the apps they use daily.