PR’s AI addiction could be a problem for the planet

AD Izzy Shipley Looks At PR’s AI Addiction And Its Environmental Cost
Last week saw a wave of agencies and brands posting sweet, funny, irreverent Barbie-doll versions of their employees, complete with matching accessories like laptops, coffee cups and pets.
The week before, there were Studio Ghibli-style versions of any image you can think of. Every Monday seems to present a new AI-generated trend for PR pros to delight in.
The beauty of AI is its accessibility – anyone can jump on a trend, or create an image that previously required the meticulous work of a graphic designer or even a costly photo shoot to create.
However, while AI is PR’s creative revolution, it has a dirty secret.
Generating a single image can consume as much energy as half of a smartphone’s battery charge, according to a recent study by Hugging Face and Carnegie Mellon University.
Given that achieving the perfect image often requires multiple prompt iterations, and agencies generate numerous creative ideas daily, the impact of PR and other creative industries on energy consumption shouldn’t be overlooked.
But the reality is that we aren’t just using AI image-generation tools to create fun images for LinkedIn, it’s increasingly becoming part of our daily work – designing campaign assets, mocking up activations, editing images, and the list goes on.
So how can we make the most of AI image-generation software while using the technology in a way that’s green?
Ask yourself – why are we doing this? Is generating this image saving you an energy-intensive photo shoot or hours of work running Photoshop? Are you doing it for real benefit to your or your client’s brand, or just jumping on a trend? While it’s easy to get swept up in the excitement of AI’s capabilities, we need to consider the environmental cost. There are so many times when AI is the obvious choice, but just as many where it should be reconsidered.
Consider whether the image is too niche. AI image software relies on source images to form its outputs and in some cases, it will naturally struggle to create an image that’s too far from what already exists, meaning that many prompts are needed in order to navigate the software closer to what you’re looking for. In some cases, AI can be more trouble (and energy) than it’s worth.
Finally, learn to write better prompts. The more detailed and specific the prompt is, the quicker the software will get to an accurate image. The more we learn about AI, the better we can work with it to use it efficiently, saving both our own time and the planet’s resources.
There’s no doubt that this software is changing our industry for the better and making new things possible. However, embracing its advantages also means navigating its downsides and considering how we marry up its use cases with our own ESG principles.
This article was originally published by PR Week on 16/04/2025