Google is pinning the blame on AI as it announces its greenhouse gas emissions have soared 48% over the past five years.

Powering its data centres, which require a surplus of energy as well as cooling methods, has placed the tech giant’s net zero gaols in jeopardy.

Based on Google’s report, global data centre electricity usage accounted for 1-1.3% of global energy demand from 2022. By using this estimate as a proxy for 2023, Google’s data centre consumption accounted for 7-10% of global data centre electricity consumption, and about 0.1% of global electricity demand.

The company’s annual environmental report blamed supply chain and data centre energy consumption as the main culprits for the metric rise in carbon emissions, saying that it now “won’t be easy” to meet its “extremely ambitious” net zero by 2030 goal.

The tech giant has invested heavily in AI, which requires data centres to train and operate their complex models such as Gemini.

Shifting to an AI-centric business model has made ambitious climate goals all the more unrealistic for tech companies putting the advancing technology over efforts to reduce and mitigate their carbon emissions.

Microsoft is also struggling, with the company’s president Brad Smith saying that their carbon negative target was now even further out of reach due to AI investment and its accompanying energy consumption.

However, technology proponents claim that AI will still help in the climate crisis – Bill Gates, co-founder of Microsoft, claimed that AI would help reduce emissions as tech leaders are willing to pay a premium for greener energy sources. While large tech companies have purchased renewable energy, the problem with AI energy consumption is much larger.

Supply chain emissions are less controllable and rely on multiple vendors, electricity still often comes from fossil fuels, and the water demand for AI data centres is also extreme.

Recommended reading

• Boards Increasingly Aware of Risks Posed by Ignoring ESG
• ESG Innovation Summit | The Two Sides of Tech’s Sustainability Impact
• This is Where it Gets Hard: A Guide to Transformative Sustainability Practices

The International Energy Agency has predicted that the energy demand of data centres could double from 2022 to 2026, amassing to 1,000 terawatt hours, which is about the same as Japan.

The director of the UK’s National Grid warned that data centre energy consumption would double by 2050, a stark reality on the consequences of AI investment.

Further, AI requires hardware to function, meaning that continued mining of precious minerals such as lithium, copper, and cobalt will continue, with potentially dire human and environmental consequences.