Egypt Edition 
UAE Edition 
Saudi Edition 
MENA <> India Corridor 
Logistics (MENA) 
Climate (MENA) 
Share
Our collective TikTok carbon footprint is larger than you might expect. How big, you ask? Slightly larger than that produced by Greece. Last year, the country as a whole emitted 51.67 mn tons of carbon dioxide equivalent (CO2e) in 2023. The platform’s total emissions, on the other hand, are estimated to be around 50 mn tons of CO2e, the Guardian reports, citing Paris-based climate strategy company Greenly.
(Tap or click the headline above to read this story with all of the links to external sources.)
On average, each TikTok user generates greenhouse emissions equivalent to driving an additional 198 km in a gasoline-powered car annually, which adds up to around 48.5 kg of CO2e per person. TikTok’s estimated emissions in the US, UK, and France alone reached 7.6 mn tons of CO2e in 2023.
TikTok’s high carbon footprint is mainly due to its addictive nature. The platform’s algorithms promote the mass consumption of videos, resulting in more minutes spent either consuming or creating content, says Greenly chief executive Alexis Normand. An average user spends 45.5 minutes daily on TikTok, but only 30.6 minutes on Instagram.
Where do TikTok emissions come from? Around 99% of the platform’s emissions come from data centers used to process and store the videos. Emissions from charging devices were also reported in Greenly’s analysis. MIT’s Technology Review records data centers in the US to be responsible for 105 mn metric tons of CO2e over the span of a year.
TikTok doesn’t disclose its emissions data to the Carbon Disclosure Project or do emissions reports for the public, unlike Google and Meta. Also, since calculations coming out of data centers don’t include emissions from smaller sources — such as commuting and offices — the estimated Tiktok emissions data are mostly an underestimation.
Sure, Tiktok’s carbon footprint is staggering, but it pales in comparison to AI’s. One AI training session can emit 284 tons of CO2e, which is five times the lifetime emissions coming out of an average passenger car, according to Nature. Meanwhile, a Statista research found that AI-powered tools like OpenAI’s GPT -3 require 2.9 to 4.6 watt-hours per request, depending on the complexity, three times as high as a simple Google search.
