A new international report highlights that integrating AI for sustainability into environmental strategies is not only viable, but profitable.
Companies that apply AI with a sustainability-focused vision are achieving significantly greater reductions in carbon dioxide (CO₂) emissions than those that still separate both areas.
The research was conducted by The Brightline Initiative, a strategic analysis center linked to the Project Management Institute (PMI).
The study was based on data from over 650 organizations worldwide, and its conclusions mark a turning point in the relationship between technology and the environment.
In this sense, it may be that in the near future, companies will seek talent with transferable skills between artificial intelligence and environmental responsibility to nurture their payrolls and reduce their carbon footprint.
## Leading companies emit less and move faster
According to the report, companies that fully integrate AI into their sustainable policies achieve an average reduction of 26% in their CO₂ emissions.
In contrast, companies that address these strategies separately barely record a timid 3% decrease.
The difference is not limited to environmental impact. Organizations with higher technological maturity also report benefits in cost savings, operational efficiency, and innovation in clean energy projects.
“AI is not a magic solution, but a driver of change,” said Pierre Le Manh, president of PMI, commenting on the results.
## Technological maturity makes the difference
The study emphasizes that the level of adoption of artificial intelligence directly influences environmental success.
Among companies using AI in an advanced manner, 31% have achieved specific goals in areas such as energy efficiency. In contrast, only 8% of those in early stages can say the same.
This finding shows that it is not just about incorporating technology, but about strategically integrating it with a long-term vision.
Organizations that manage to connect data, leadership, and collaboration between areas achieve better results and greater environmental impact.
## A three-stage path to integrating AI and sustainability
The report proposes a simple yet powerful action model for those who have not yet taken the step:
– **Strengthen the database:** Gather reliable information to make precise decisions.
– **Train cross-functional teams:** Unite technology, operations, and sustainability on the same axis of work.
– **Link strategy with senior management:** Make sustainability a central objective in decision-making.
The gap between leaders and laggards is widening, and those who do not adapt run the risk of losing competitiveness in a market that increasingly values real sustainability.
## The flip side
Although AI can accelerate corporate sustainability, its current use also has an environmental cost that should not be overlooked.
The training of advanced models consumes large amounts of electricity and generates significant carbon emissions.
A revealing example is the GPT-3 model, whose training process used 1287 MWh of energy and emitted 550 tons of CO₂, a figure equivalent to 33 flights from Australia to the UK, according to data from a study by the specialized portal ScienceDirect.
But the impact does not end there. Each interaction with tools like ChatGPT also requires energy. As the use of these systems grows, their environmental footprint also multiplies.
Experts estimate that AI could account for more than 30% of global energy consumption by 2030 if efficiency measures are not adopted.
## Green AI: a proposal to reduce the technological footprint of AI
In response to the high energy consumption of artificial intelligence systems, researchers and developers have begun to promote a new trend called Green AI.
This approach proposes to make AI not only useful for combating climate change but also more efficient and sustainable in itself.
The concept revolves around two complementary strategies. On one hand, “Green-by AI” promotes the use of intelligent systems to optimize industrial processes, reduce emissions, and support environmental policies.
On the other hand, “Green-in AI” aims to improve AI models themselves, minimizing their energy consumption and reducing emissions resulting from their training and operation.
Both strategies seek to mitigate the environmental impact of the growing use of this technology without hindering its progress.
## Controlling the environmental damage of these technologies
Specialists have proposed several strategies to make AI less polluting:
– **Algorithm optimization:** Techniques such as neural network pruning or model distillation allow performance to be maintained while reducing energy consumption.
– **Use of specialized processors:** Tensor Processing Units (TPUs) consume less energy than traditional GPUs in many AI tasks.
– **More sustainable data centers:** Migrating to the cloud and opting for renewable energy significantly reduces the carbon footprint.
This approach is already beginning to be reflected in the development of models like DeepSeek, which adopts energy-efficient principles from the design stage.
## Regulation, a key factor in the future of sustainable AI
Beyond technical innovation, regulation also plays a decisive role. In the European Union, the AI Act proposes sustainability criteria for high-impact models and promotes codes of conduct in energy efficiency.
In the United States and China, policies do not yet impose environmental limits on AI. However, some companies have started taking voluntary measures such as using renewable energy and improving their training processes.
Faced with the rapid advancement of technology, experts agree that the development of global regulations will be essential to balance progress and environmental responsibility.
