For a majority of marketers, AI will play a significant role in their plans over the next 12 months, but such a powerful tool is also resource hungry – developing a robust strategy will require a balance of competing objectives to make AI effective over the long term.
This is a key theme in The Marketer’s Toolkit 2025.
One ChatGPT request, for instance, takes up to 10x more energy than a query on Google’s search engine. On its part, Alphabet, Google’s owner, has said AI was central to its emissions increasing by 48% from 2019-23 – and will make cutting its future emissions hard.
Given marketers’ ambitions for AI, its usage threatens to pull them in the opposite direction from the existential problem of prioritising sustainability in a time of climate crisis.
What’s going on
Emissions from algorithmic advertising uses aren’t new: most of the emissions generated by programmatic buying arise from ad selection. But AI goes many steps further, with one generated image requiring as much energy as charging a smartphone.
Ignoring this issue may ultimately invite criticism from eco-conscious consumers and unwanted regulatory scrutiny.
There are steps that marketers can take, however, though some are more convincing than others. Reducing ad waste, eco-conscious media investment, and right-sizing LLMs to seek the best tool for the job rather than the most powerful.
