Across the telecoms sector, operators, vendors and other suppliers are pushing towards renewable energy and sustainability goals with an uncommon unity. Given that operating a network requires a constant supply of power, it makes sense that to pivot towards cheaper and more sustainable sources of energy – but behind the talk, are companies truly taking action on their Net Zero goals?
The GSMA’s Mobile Net Zero Report 2026 examined industry progress on the broad goal of achieving Net Zero by 2050 and found that increased access to renewable energy is fundamental to achieving this. We spoke to Steven Moore, Head of Climate Action at the GSMA, to find out where progress is being made, and what more needs to be done.
The report mentions that a 45% reduction is required on the science-based pathway, but it’s looking like the sector might fall short of that by 2030. Why is the industry not on track? Further progress in Asia particularly is going to be essential to accelerate that – why is Asia so significant for this?
The steepness of the reduction curve is aligned to limiting heating to 1.5 degrees – effectively it means a 45% cut by 2030. Beyond that to 2040, there’s a project underway with the ITU to work out what the next stage of the science-based pathway will be, but the end goal needs to be Net Zero by 2050. We’ve had a 13% cut from 2019 to 2024, and that’s despite seeing a very big increase in both data traffic across networks and mobile connections. We’ve managed an overall global decrease despite effectively serving more customers and providing a greater service, but we are short – at about a 5% reduction in 2024, which is the last full year that we have data for. We need to be about a 7.5% annual reduction by 2030, so to hit that there needs to be an acceleration. There are two very significant regions to help us achieve that – they’re both in the same part of the world, but we separate them in the GSMA: the Greater China region and APAC more broadly. In terms of the number of connections and the electricity use across those two areas, the overall Scope 1 & 2 carbon emissions are really significant. What happens in those markets will determine whether we’re able to hit that 2030 global target. China has made fantastic progress on renewables; they’re adding more renewable capacity than the rest of the world combined each year, but still are also using a significant amount of coal to power electricity, and we’ve seen significant increases in electricity use by Chinese operators over the last few years. We have seen a drop; Chinese operators have started to purchase some renewable energy, so that’s helped, but the world is really watching China to see what happens to its emissions over the next few years. Overall they’re stabilising, potentially reducing, but then we’ve had this challenge with the energy crisis in terms of receiving things like oil and gas; is that going to mean that there’s more coal that’s used? The APAC region more broadly is also an interesting one because mobile connections are continuing to grow, whereas they’ve plateaued more in Europe and North America and other parts of the world. We also highlight in the report the potential growing investments in AI; that’s obviously quite uncertain.
AI’s power consumption is a controversy in itself – how is this impacting the industry’s Net Zero goals?
Some operators have made announcements about investments in AI; the very largest tech companies have made some extremely large announcements in terms of their financial investments into AI. There is a question mark around how much of this will actually end up being built out because of challenges around the chips and memory shortage, which we know is impacting the cost of mobile phones as well, and whether they can get sufficient power for these new data centres – that’s the other potential constraint. We’ve looked at what’s been announced by some of the biggest companies, and how telco announcements compare against those. Those are some of the headwinds to reductions over the coming years, but there’s significant uncertainty around those. We also talk about continued improvements in energy efficiency; we know that the rollout of 5G has helped, that switching from copper to fibre is also helping, and that switching off 2G and 3G legacy networks seems to support reducing network emissions in countries where they’ve been shuttered. We’re in the middle of developing 6G and we’re having lots of discussions around the sustainability of this. There’s also been an increase in renewable use across the sector – it was around 10% in 2019, now around a quarter of electricity used is directly purchased by operators. This is in addition to renewables that are already on grid worldwide, which operators get automatically by buying electricity from a country’s grid, but lots of operators go beyond this with specific investments – either buying renewable energy certificates or entering long term power purchase agreements. Operators are a good match for renewable energy providers as they want longer-term customers; operators can say that they need energy for the next 20 to 25 years, and these long-term purchase agreements have enabled the construction of several new renewable energy facilities worldwide because of the guarantees around the buying. We also see a lot of on-site renewables, particularly in off-grid locations or areas with a poor grid. This is challenging as there’s often a lack of available space; masts are in rented locations, and this restricts the amount of onsite renewable energy we’re able to generate as an industry. That’s never going to power more than a couple of per cent in terms of overall electricity demands, but where we can deploy it it’s essential.
The role of tower companies is particularly significant to emerging markets – the ownership structures are a bit different to mobile network operators, which in general are more publicly listed companies so there’s more transparency around their emissions, with annual sustainability reports. This is not generally the case with towercos – some of them publish their results, and are making good progress on reducing Scope 1, 2 and 3 emissions, Cellnex being a good example – but there are many where it’s less clear. We included some estimations of the amount of diesel used by tower companies in 2024 – around 2 billion litres – which is obviously a significant cost, and there are major environmental implications of the carbon emissions of 2 billion litres of diesel being burnt. We know that the 100 largest tower companies cover about 4 million sites around the world, so they’re a significant part of the market. What we also explored in the report was how well aligned the lease agreements are, i.e. the contracts between tower companies and mobile network operators, in terms of supporting some of the targets. 81 operators now have near-term science-based targets, but we see a smaller proportion of tower companies that have these targets as well. That’s a challenge for mobile network operators if the company that they’re contracting isn’t as aligned as they are in terms of reducing emissions and switching to renewables.
Operators have a degree of leverage here, but that’s largely going to be through how agreements are structured. Is there much of an appetite for gearing business models towards those that encourage decarbonisation among tower companies? The efforts of operators to reduce their emissions obviously are apparent, but if tower companies are not doing enough, the impact is going to be diminished.
That’s the big question mark: what are the different contract structures used across the industry, and are they supporting decarbonisation? That’s what we provide analysis around – we give examples of which are likely to work in the interests of the operators for reducing emissions, and others in which there’s not really any incentive for the tower company to reduce emissions. We highlight it as an issue across the industry, and explain how incentives might be better aligned – pointing to contracts which would work more in the interest of operators with targets, to say to them: “if you’re looking to renegotiate or when the contract comes to an end, you might want to consider restructuring it in a slightly different way, so that you and the tower company are on the same page”. We’re interested in engaging with our companies more: we’ve certainly spoken to them from the climate side. We want to send signals that this is where operators are, this is where the industry as a whole is moving to and see if we can have more of them get on board and support reduction efforts. We would love to see some more transparency around reporting of emissions because it helps us. We do this report every year, and there’s a data gap, so it makes it more difficult for us to understand where exactly emissions are going if we’re having to estimate and infer some of these figures.
Operators are inherently consumer focused; they feel that pressure but also a genuine concern and drive to pursue this. A lot of the emissions from the sector are Scope 3, i.e. supply chains, which encompasses towercos, so perhaps feel less pressure to be seen to be decarbonising or reducing emissions. How can that be addressed? How can we engage that side of the sector in order to start driving down carbon emissions across Scope 3?
This is this is one of the key challenges, particularly around Scope 3, but around how companies share these targets internally and make sure the organisation is moving in the same direction on them, because climate targets are so far-reaching they now are intersecting right across the organisation. Scope 1 & 2 emissions are focused on network operations, but Scope 3 is all about procurement, and operators obviously procure lots of different things, not just energy. There are conversations to have with device manufacturers and network equipment suppliers, and there are all sorts of other things that they might buy as well. We’re seeing some operators moving into selling other types of equipment, so having your procurement team fully up to speed in terms of what the climate targets are, and then engaging your suppliers, having that conversation to say, « We’re aiming for this reduction by 2030. Are you measuring your emissions? How are you managing them? Have you set your own science-based targets? Are you cascading that back into your supply chain?” Those are all fundamental questions that we see the leading operators [asking] and have quite a positive relationship with their suppliers around this, because they’re sending a signal of the direction they want to head in, but they’re also setting out a reasonable time frame for suppliers to respond and take action. We have seen some, particularly European operators, say to their suppliers, « If you don’t meet these requirements, then you effectively are deselected.” So there’s an incentive for them to act, and operators provide support, but there’s also there’s a hard backstop that if they don’t see movement, then they’ll look to choose others. That’s on the procurement side, but we see climate action targets cutting across into finance; when you’re making the business case for a particular switch in approach, you might be looking to raise money as well. We’ve seen green bonds be used quite extensively by operators where they want to make sustainable investments.
So transition planning needs to be fairly holistic even if procurement is one of the main pitfalls?
Overall, something that we worked on from the GSMA side that we published last year was guidance on climate transition planning. When you’re reducing emissions, you need to take a whole company approach to it – it doesn’t really work if you just have a sustainability team trying to work on this in isolation. You need to have all different departments across the company brought into it, and each of them taking responsibility for the area of the target that they’re managing. We see reducing emission size, the mitigation part, being just as important as the adaptation piece as well. Both of those form a climate transition plan: how are you reducing emissions? Obviously we’re trying to avoid changing the climate much more than we have done already, and then accepting that there is already a significant amount of climate change built in. So, what are operators doing to address that, and how can they make their networks more resilient? That’s also an area that we’ve been working on more with operators over the last few months to try and understand how are they addressing some of the bigger climate challenges that are now confronting them? We’ve seen enormously extreme weather events over recent years. The mudslide through Valencia, the flooding of the Rhine in Europe, the U.S. in terms of wildfires, which are now becoming much more common in Europe. A third of Pakistan was underwater a couple of years ago. Southeast Asia is continually battered by typhoons. For operators, it’s business as usual in terms of keeping a network running, but there are now climate events, weather events that were completely unexpected. Portugal had its first very strong storm at the beginning of this year, with winds in excess of 80 to 100 miles an hour. That wasn’t even part of their risk register because it had never happened before. That’s also something else that we touch on in the report, looking at how the industry is addressing this challenge and what some of the potential costs might be of future damage around this, because we can see that’s likely to increase in the future.
What’s the impact of data centres? Are we only factoring in operator data centres in terms of the emissions? What about non-operator data centres?
We’ve looked at what the hyperscalers have announced and the type of energy use that will entail. In comparison, telecom operators are a small fraction. Geographically, the Middle East and APAC regions are more likely to be building AI data centres – Japan, South Korea, Malaysia, Singapore, although it has some energy challenges, and the Middle East. We’re not seeing it so much from operators in Europe, North and Latin America. Some operators are building AI data centres but not to the same scale. Critically, the scale of telecom operator AI data centres is much smaller than hyperscalers. Given that 81 operators are aiming for science-based targets where they’re building data centres, it’s much more likely to be powered by renewable energy as they want to hit these targets. We will evaluate this over the next few years, seeing how data centre buildout will affect our energy usage as a sector, and how many are being powered by renewable energy or other sources. One of the benefits of powering them with renewables is that they’re actually very cost competitive, particularly in certain parts of the world, and a renewable energy installation – a wind farm or a solar farm – can be quick to construct, which is a challenge at the moment in terms of how quickly can you plug into a new energy source or if there’s available energy on the grid, so this seems to be one of the biggest constraints in terms of building these new data centres. Given some of the memory and power challenges, as well as planning constraints, we can see policymakers are hesitant. We’ve seen local pushback, particularly in the U.S. where very big installations have been built. Local electricity prices have gone up, so we’re starting to see opposition to building data centres. So, quite significant uncertainty around announcement versus what will be the actual build out in the future.
We’ve seen pushback in terms of the ability to connect to the grid and the reliability, but also the incentives that have been offered to build out data centres – countries are realising they don’t necessarily have the ability to meet demand. The report notes that in terms of renewable energy powering data centres, operators are consuming about 300 terawatt hours of electricity in 2024, of which 70 was renewable, so less than a third of the total usage.
That’s of purchased renewables. I should clarify because it’s tricky to come up with an overall figure because if you look at electricity produced globally, around a third of it on grids is from renewable sources – wind and solar, also hydro power. We add that figure to the purchased one, but there is a bit of overlap so the accounting becomes tricky. We specify the amount that’s purchased directly by operators, but there will be an addition of what’s already on the grid. Operators in general want to use more renewable energy; the industry is a very eager purchaser of renewables, and that is quite a challenge because many markets around the world don’t have liberalised energy markets. They might have one single provider which might offer you a green tariff that is more expensive than the normal tariff, whereas in a liberalised energy market like the UK and other European markets, you can directly contract with a renewable energy generator, and you can enter into a long-term power purchase agreement. One of the recommendations for policymakers and regulators is to improve access to renewable energy for mobile network operators. We need a fully liberalised energy market because we need to have wheeling in the technical part – wheeling of electricity across the grid, because we have many sites across the country, unlike a data centre; you can build a solar farm next to the data centre, plug it in, and then you’re powering that data centre from renewables. We can’t do that because most of our electricity flows through the grid, so we need to have a system in which we can match our electricity demand through a market design, and tell an energy generator “we use this amount of energy, we’re going to purchase it from a renewable energy generator, and they can provide that to us.” These are technical conversations that we’re having with policymakers and regulators in certain countries. We’ve had some really positive discussions with the Malaysian government over recent years, and they are moving in that direction and making it easier for corporates to purchase renewable energy. But that is a significant challenge for us in quite a few markets, and particularly developing markets.
Your statistics show that a lot of renewables are purchased in Europe and the US, but in Latin America the figure is 45% – is it lower in Asia and Africa? I imagine the energy markets there are not as liberalised as would be ideal in this scenario.
No, and in APAC, there’s a challenge around one single monopoly provider. There also is that situation in Africa, but there’s the added challenge of a lack of a grid in many cases. We see so many off-grid sites, and obviously that’s where we see a significant amount of diesel use. Or if there is a grid, it’s not reliable enough, so you need to have a diesel generator back up. We’re very fortunate in that recent years there’s been a huge amount of technological development in solar panels, and their efficiency, and then also battery development in general in terms of storage, and also the cost of batteries. That’s all very helpful in terms of looking at alternatives for diesel generators. Fuel cells are also another option that I know some operators are exploring as well; there are still some challenges around security though, in that where they have been installed, there’s been theft associated with some of them because they’re seen as quite valuable assets. There are things that operators want to do; they’re moving in that direction, but there are some challenges in order to achieve that.
