New Research Reveals Awareness of AI’s Environmental Impact
A fascinating new study from Hiyield explores the environmental impact of artificial intelligence, offering valuable insights for organisations navigating the delicate balance between innovation and sustainability.
The Research at a Glance
Hiyield’s comprehensive study, combining survey data from 43 business decision-makers with University of Exeter research, reveals some striking findings about AI’s environmental footprint:
Trust and Implementation
Interestingly, despite AI’s widespread adoption, trust remains a significant challenge.
The study found that less than a fifth (19%) of respondents always trust AI outputs, while 16% never trust them. The majority (65%) take a cautious approach, only sometimes trusting AI results.
The Environmental Challenge
The research highlights three primary concerns from business leaders:
- Lack of regulation around AI’s carbon footprint
- High energy consumption often generating low-value outputs
- Rushed adoption without proper environmental impact assessment
“Companies are doing it because they can do it, but not considering if they should do it,” noted one survey respondent, highlighting a growing need for more thoughtful AI implementation.
Matt Ville, Founder and CTO of Hiyield, said:
While AI presents significant challenges, it also offers immense potential to transform businesses and daily life, shaping a more sustainable future. At Hiyield, our partnership with Digital Carbon Online enables us to measure and monitor the digital carbon footprint of our own website, as well as the apps and sites we create for our clients. Collaborations like this are key to driving responsible, sustainable innovation forward.
The Way Forward
While AI poses environmental challenges, it also offers potential solutions for sustainability:
Hiyield’s research outlines five key recommendations:
- Integrate green technology (renewable-powered hosting, energy-efficient data centres)
- Develop a clear AI strategy aligned with sustainability goals
- Focus on high-impact applications
- Regular carbon impact assessment
- Collaborate with providers on energy optimisation
Commenting on the report, Scott (CEO, Digital Carbon Online), said:
This report is a fascinating glimpse at both the increasing usage of Large Language Model (LLM) / Generative AI across business, as well as the growing concern of its environmental impact. I’d be keen to discover whether these findings hold true across a broader spectrum of responders and how opinions differ comparing LLMs against the benefits/consequences of other AI technologies such as Computer Vision, Reinforcement Learning, Explainable AI (XAI).
Our mission at Digital Carbon Online is for a more sustainable digital future, so the role of AI is an important question. We see numerous positive benefits of AI, but there are also many downsides that need understanding and addressing. On top of all that, not all AI have the same balance between good and bad, so it really is a situation of picking “the right tool for the job”. Being able to make that judgement call is difficult because we simply don’t have all the data – reports like this are important steps to getting there.
