The digital It is now an essential part of our lives: it allows us to work, communicate, inform ourselves and entertain ourselves in ways that were unthinkable until a few years ago.
However, despite the fact that it seems weightless and free of matter, the digital world brings with it very real environmental consequences, but hidden!
High energy consumption, the resulting emissions and the exploitation of natural resources are just some of the main environmental impacts that the digital world causes.
With the advent ofgenerative artificial intelligence, this phenomenon has also amplified exponentially!
Train and use models increasingly powerful requires enormous amounts of electricity, water and materials, making the digital sector one of the protagonists of environmental challenges of our time.
In this article we will analyze what the digital carbon footprint is, what are the concrete impacts of digital and we will see the 5 main actions that are being tried to take to reduce the digital footprint and contribute to a more sustainable development.
{summary#bullet-1}
What is the digital carbon footprint
The digitalization is now an integral part of our daily lives and business processes: from the use of smartphones and computers to the development of advanced technologies such as artificial intelligence.
This progress brings with it numerous advantages Like:
- optimization of processes;
- reduction of waste;
- decrease in the use of paper.
However, behind every digital service there is also a least visible side: an increasing environmental impact!
With the term Digital Carbon footprint (digital carbon footprint) indicates the sum of the greenhouse gas emissions generated along theentire life cycle of digital technologies. This is not only about the energy needed to power devices, servers and communication networks, but also about resources consumed And of emissions produced during the production, transport and disposal of the hardware.
Every email sent, every streaming video or every trained artificial intelligence algorithm leaves behind a trace of equivalent CO2.
This imprint therefore includes different sources of emission:
- the use of devices such as computers and smartphones, which require electricity;
- digital infrastructures, in particular data centers, real “energy hearts” of the web;
- the production and the end of life of the hardware, with industrial processes that involve the consumption of materials, transport and management of electronic waste.
Understanding the digital carbon footprint is essential for deal with it in a conscious way the environmental challenges related to the technology sector.
While on the one hand digital can be an ally in the ecological transition, on the other hand it is important recognize its impact and work to reduce it, especially in a context in which the spread of artificial intelligence risks further amplifying its emissions.
{summary#bullet-2}
What is the digital environmental impact of people in everyday life
When we think aboutenvironmental impact of digital, we often only imagine the large data centers of multinational technology companies. In reality, even ours daily actions online have a concrete environmental cost, often underestimated.
An example? The generative artificial intelligence.
Several studies highlight how a single question addressed to systems such as Chat GPT It involves, on average, the emission of 4.32 grams of CO₂ and a consumption of water that, in five interactions, reaches approximately 500 ml, the equivalent of a small bottle.
If we compare this data with traditional research on Google, the difference seems obvious!
A search engine query emits about 0.2 grams of CO₂, while five searches consume on average 46 ml of water.
Numbers that, taken individually, may seem small, but if multiplied by millions of requests generated every day they become a far from marginal environmental footprint.
This brings us to a Important reflection!
In an increasingly digitized world, where digital tools accompany us in every aspect of personal and professional life, it is essential to ask ourselves how necessary the services we use are really and how we can make them a More conscious use.
Digital technology has enormous potential to promote sustainability, but to make this happen it is also necessary to be aware of its invisible costs.
{summary#bullet-3}
What are the impacts of the digital sector
The digital sector is today one of main engines of economic and social transformation, but its growth involves environmental consequences increasingly significant.
From manufacturing devices to powering data centers, every component of the digital ecosystem leaves an imprint that goes beyond simple energy consumption.
Environmental impact of data centers
I Data Center they are the central buttons of the digital world. These important digital infrastructures are responsible for:
- keep the data;
- power the cloud;
- support streaming;
- allow the development of increasingly advanced technologies such as artificial intelligence.
This enormous invisible infrastructure, however, has a significant environmental cost!
The first to take into consideration is the energy consumption. In 2021, data centers were responsible for about 1-1.5% of the world's electricity, equal to more than 220 TWh. A share that, with the spread of increasingly energy-intensive digital services, could represent 20% by 2030.
According to studies by the University of Berkeley, only for the training of GPT-3, for example, 1,300 MWh of electricity were needed: the equivalent of annual consumption of an entire neighborhood.
The more AI models grow, the more the consumption curve accelerates relentlessly!
What weighs not only the energy required by the servers, but also that necessary for their cooling: in fact, up to 50% of the total consumption of a data center derives from air conditioning systems.
The combined effect of these factors results in greenhouse gas emissions increasingly relevant. Today, data centers account for about 0.5% of global CO₂ emissions, but estimates speak of a possible growth of up to 1% by 2030.
Finally, we must not overlook the impacts related to territory and materials where plant construction causes:
- the occupation of large areas of land;
- high land consumption;
- pressure on ecosystems.
This growing energy pressure raises a critical point: without structural interventions and a rapid transition to renewable sources, the expansion of data centers could slow down the abandonment of fossil fuels and jeopardize the achievement of global decarbonization goals.
Water impact
The digital sector, and in particular artificial intelligence and the immense infrastructure that supports it, do not consume only energy: their effect on the environment also extends to water resources, often invisible but just as crucial.
Also according to a study by the University of Berkeley, the training of GPT-3 It took about 700,000 liters of fresh water: a quantity sufficient for 17,500 showers.
In addition, according to a prospective estimate, by 2027, artificial intelligence could result in a global water collection of between 4.2 and 6.6 billion cubic meters.
Roughly as if theThe UK's entire annual water consumption would be used to make AI work.
This pressure on resources is manifested in a concrete way in the territories:
- data centers, essential for powering models, are often located in areas already subject to water stress;
- new infrastructures are planned in regions such as Texas, Arizona, India and Saudi Arabia, where the availability of resources is already limited, increasing the risk of conflicts with agricultural or urban communities.
Despite this critical issue, there are signs of change. The industry is experimenting waterless cooling systems, such as “waterless” technologies or closed circuits, already adopted, for example, by 64% of commercial data centers in England.
At the same time, the digital sector is engaging in the use of AI to improve water management: intelligent systems are being developed in Europe, from IoT sensors to real-time data analysis, which make water services more efficient and sustainable, for the benefit of both the environment and communities.
Mining impact
Behind the invisible efficiency of digital tools hides a production chain that is based on hard-to-find minerals which:
- lithium;
- cobalt;
- rare earths;
- nickel;
- copper;
- silicon.
These materials, essential to power the digital world, are extracted through processes that cause serious environmental and social consequences.
The extraction of the lithium, for example, involves the removal of enormous quantities of minerals that cause deforestation, groundwater contamination and land degradation.
The cobalt, extracted mainly in the Democratic Republic of the Congo, is associated with phenomena of social inequality, exploitation and serious risks to human rights, including forms of child labor.
Le rare earths, used for the most sophisticated components of digital hardware, are extracted in regions with poor environmental guarantees: a kilogram of these can produce up to 2 tons of toxic waste, including radioactive waste.
More generally, the mining activities cause:
- erosion;
- loss of biodiversity;
- water and soil pollution;
- release of chemicals;
- profound changes in natural landscapes.
And yet, it is still possible to reverse the course!
The practices of responsible extraction, such as Direct Lithium Extraction (DLE), promise a reduction in water and territorial impact, while the growth of mineral recycling, which today covers only a small part of demand, could alleviate mining pressure and promote greater circular economy.
{summary#bullet-4}
5 ways to reduce your digital environmental impact
Reduce the environmental footprint of digital and artificial intelligence is not only a technological challenge, but a collective responsibility which involves companies, institutions and citizens. Here are 5 concrete actions to contribute to a more sustainable digital future.
1. Develop more efficient AI models
Research must be oriented towards algorithms capable of offering high performance while consuming less energy and water. Efficiency isn't just a technical issue: it's the key to making AI a tool for progress without burdening the environment excessively.
2. Power data centers with renewables
Data centers are the heart of digital technology and their footprint depends largely on the energy source used. Powering them with renewable sources and adopting more sustainable cooling systems means drastically reducing emissions and water impact.
3. Optimize data management
Keeping only the necessary information, reducing the duplication of content and improving the efficiency of digital flows are practices that make it possible to limit energy and resource waste. A leaner 'digital diet' is beneficial both for the environment and for companies.
4. Promote conscious use at an individual level
Every person can contribute with daily choices: from using AI only when really necessary, to extending the life cycle of PCs and smartphones with proper maintenance. Even simple practices such as periodically cleaning up useless files and data or reducing the number and weight of emails sent make it possible to contain consumption and emissions. The adoption of these behaviors on a large scale multiplies their positive impact.
5. Establish transparent rules and standards
To ensure that technological progress is truly sustainable, clear rules are needed: consumption reporting obligations, independent audits and shared standards for evaluating environmental impact. Transparent governance is the basis for building trust and accountability in the sector.
