AI Drives Up Microsoft Emissions: 2030 Climate Goal at Risk

The Context: An Ambitious Climate Goal

In 2020, Microsoft committed to being a "carbon negative" company by 2030, meaning it would remove more carbon from the atmosphere than it emits. To achieve this, it implemented energy efficiency measures, renewable energy, and offset programs. However, the rise of artificial intelligence has drastically changed the landscape. This commitment is part of a broader trend: since 2019, major tech companies like Google, Apple, and Amazon have also announced ambitious climate goals, but operational reality has shown that the expansion of AI is straining those plans.

What Happened?

According to Microsoft's 2026 environmental report (covering fiscal year 2025), total net emissions rose from 16 million to 20 million tons of CO2 equivalent, a 25% increase from the previous year. Gross emissions reached 34 million tons, partially offset by carbon credits. The company now emits as much CO2 as entire countries like Panama or Lithuania, according to GeekWire. This increase contrasts with the previous trajectory: between 2020 and 2023, Microsoft had managed to reduce its net emissions by 6%, but the emergence of generative AI reversed the trend.

Melanie Nakagawa, Microsoft's chief sustainability officer, told The Verge that the company remains committed to its 2030 goal but acknowledged that the expansion of data centers for AI is the main factor behind the increase. The computing demand for training and running large language models (LLMs) requires enormous amounts of electricity, which drives up the carbon footprint. According to Xataka, the report details that Scope 1 (direct) and Scope 2 (purchased electricity) emissions grew by 30%, while Scope 3 (supply chain) emissions increased by 22%. The latter is relevant because it includes the manufacturing of specialized hardware, such as NVIDIA GPUs, whose production process is carbon-intensive.

Why Is This Important?

Microsoft is not an isolated case. Google, Amazon, and Meta have also seen their emissions rise due to AI. Google, for example, reported a 48% increase in its emissions in 2024 compared to 2019, mainly due to the energy consumption of its data centers. This phenomenon calls into question the sustainability of the current race to dominate artificial intelligence. If major tech companies fail to decarbonize their infrastructure, AI could become an obstacle to global climate goals.

Furthermore, Microsoft's case illustrates the tension between technological innovation and environmental responsibility. The company has invested billions in AI, including its partnership with OpenAI and the development of its own Azure Maia chips. However, the environmental cost could erode its reputation and expose it to criticism from investors and regulators. In 2024, a group of Microsoft shareholders filed a resolution demanding greater transparency in measuring emissions associated with AI, noting that the company was not adequately reporting the impact of its technology partners.

Consequences for the Sector

• Regulatory pressure: Governments and international bodies could tighten emission standards for data centers. The European Union is already considering including data centers in its emissions trading system, and in the United States, the EPA has proposed new guidelines for high-energy-consuming facilities.
• Investment in clean energy: Microsoft and other companies will accelerate their investments in nuclear, solar, and wind energy to power their data centers. Microsoft has already signed power purchase agreements (PPAs) for over 10 GW of renewable energy and is exploring small modular reactors (SMRs) for 2027. Amazon, meanwhile, has acquired a nuclear data center in Pennsylvania.
• Hardware innovation: More efficient processors will be sought, such as custom AI chips (TPUs, Maia) and the integration of liquid cooling technologies. Companies like NVIDIA have announced that their next GPUs will reduce energy consumption per inference by 40%.
• Carbon offsets: Demand for carbon credits will increase, but their effectiveness remains controversial. Microsoft has invested in direct air capture (DAC) projects, such as Climeworks in Iceland, but these are still costly and limited in scale.
• Transparency and standards: Companies are expected to adopt more rigorous reporting frameworks, such as the GHG Protocol, and certifications for "green AI" for cloud services may emerge.

What Should Readers Know?

Microsoft's dilemma is not just its own; it is the dilemma of the entire tech industry. AI promises enormous advances, but its ecological footprint is real and growing. Consumers and businesses adopting AI services must be aware of this hidden cost. On the other hand, AI itself can help optimize energy consumption and accelerate the energy transition, creating a virtuous circle if managed well. For example, Google has used AI to reduce cooling consumption in its data centers by 40%, and Microsoft has developed models that predict energy demand to adjust workload.

In summary, Microsoft's failure to meet its climate goal is a wake-up call. The most transformative technology of our era cannot be developed at the expense of the planet. Solutions exist, but they require political will, investment, and above all, transparency. The question remains whether tech companies can reconcile their AI ambitions with their environmental responsibility before the climate takes its toll.