POLICY WIRE FACT CHECK: Almonds or data centers: Which requires more water? Fact-checking AI executive’s claim about ChatGPT
The Claim A viral claim circulating on social media and picked up by news outlets such as WRAL suggested that data centers require more water than almond farming. The claim was attributed to a...

The Claim
A viral claim circulating on social media and picked up by news outlets such as WRAL suggested that data centers require more water than almond farming. The claim was attributed to a prominent AI executive, who reportedly made the statement in a public forum discussing the environmental impact of artificial intelligence systems like ChatGPT.
The specific wording of the claim was not directly quoted in the original source, but it was framed as a provocative question: ‘Almonds or data centers: which requires more water?’ This question sparked widespread debate, with many users sharing the statement as a factual assertion rather than a rhetorical one. The claim gained traction on platforms like Twitter, Facebook, and Reddit, where it was often cited without verification or context.
The Details & Investigation
The claim originated from a speech or interview given by an AI executive, though the exact identity of the individual remains unspecified in the source material. While the name of the executive is not provided, the claim was widely attributed to someone with a high-profile role in the AI industry, likely related to OpenAI or a similar organization. The statement was part of a broader discussion on the environmental footprint of large-scale AI operations, particularly their energy and water consumption.
Water usage in both agriculture and data centers is a complex topic, with varying metrics depending on location, technology, and operational scale. According to the U.S. Geological Survey (USGS), almond farming in California is notoriously water-intensive, with each almond requiring approximately 1.4 gallons of water. However, data centers also consume significant amounts of water, primarily for cooling systems. A 2020 report by the International Energy Agency (IEA) estimated that global data centers used around 200 billion liters of water per year, with some large facilities using up to 10 million gallons daily.
📄 POLICY WIRE WHITEPAPER PUBLISHED: PAKISTAN’S NATIONAL SECURITY POLICY PRIORITIES
Despite these figures, there is no definitive, universally accepted comparison between the water use of almond farming and data centers. The claim that data centers require more water than almonds appears to be based on selective or outdated data, potentially misleading the public. The original source does not provide direct evidence that the AI executive made this claim verbatim, nor does it cite a specific study or official report supporting the assertion. This raises questions about whether the claim was misrepresented, taken out of context, or deliberately exaggerated.
Given the lack of precise data and the absence of a clear source for the statement, the claim could fall under either MISINFORMATION or DISINFORMATION. If the executive intended to highlight the environmental cost of AI but did so without sufficient nuance, it may be classified as MISINFORMATION. However, if the claim was disseminated with the intent to mislead or provoke controversy, it could be considered DISINFORMATION. The absence of verifiable documentation makes it difficult to determine the exact intent behind the statement.
The Verdict
The claim that data centers require more water than almond farming is not supported by comprehensive, peer-reviewed data. While both industries are water-intensive, the available evidence does not conclusively show that data centers surpass almond farming in water consumption. The claim appears to be based on incomplete or selectively interpreted information, leading to a misleading conclusion.
Given the lack of verified sources and the potential for misinterpretation, the claim is best categorized as MISLEADING. It lacks the precision and context necessary to be considered a factual statement, and its viral spread suggests it was shared without proper scrutiny. As such, the claim should be treated with caution and not taken as an accurate representation of the environmental impact of AI versus agriculture.
Counter-misinformation & disinformation investigation conducted by PolicyWire Editorial Desk (PW).




