Associated Incidents
At first glance, there was nothing out of the ordinary about Saturday's wind forecast for Camas Prairie, Idaho.
"Hold onto your hats!" said a social media post from the local weather office in Missoula, Montana. "Orangeotild" had a 10 percent chance of high winds, while just south, "Whata Bod" would be spared larger gusts.
The problem? Neither of those places exist.
Nor do a handful of the other spots marked on the National Weather Service's forecast graphic, riddled with spelling and geographical errors that the agency confirmed were linked to the use of generative AI.
The blunder --- not the first of its kind to be posted by the NWS in the past year --- comes as the agency experiments with a wide range of AI uses, from advanced forecasting to graphic design. Experts worry that without properly trained officials, mistakes could erode trust in the agency and the technology.
NWS said AI is not commonly used for public-facing content, nor is its use prohibited. The agency said it is exploring ways to employ AI to inform the public and acknowledged mistakes have been made.
"Recently, a local office used AI to create a base map to display forecast information, however the map inadvertently displayed illegible city names," said NWS spokeswoman Erica Grow Cei. "The map was quickly corrected and updated social media posts were distributed."
A post with the inaccurate map was deleted Monday, the same day The Washington Post contacted officials with questions about the image.
Cei added that "NWS is exploring strategic ways to continue optimizing our service delivery for Americans, including the implementation of AI where it makes sense. NWS will continue to carefully evaluate results in cases where AI is implemented to ensure accuracy and efficiency, and will discontinue use in scenarios where AI is not effective."
A Nov. 25 tweet out of the Rapid City, South Dakota, office also had misspelled locations and the Google Gemini logo in its forecast.
NWS did not confirm whether the Rapid City image was made with generative AI.
The Weather Service often creates experimental forecasting products, said John Sokich, who worked there for 45 years before retiring in January 2025. "But it's their policy to ensure that type of experimentation is labeled. They have a rigorous process for testing a product before it becomes operational."
Sokich added that he thought not labeling these images as experimental was "just an oversight."
In the past year, hundreds of employees have been fired, retired or left NWS as part of the Trump administration's efforts to scale down the federal government. The National Oceanic and Atmospheric Administration --- NWS's parent agency --- has vowed to refill many of those positions by September, but the losses have strained the service and its employees.
"If there's a way to use AI to fill that gap, I'm not one to judge," said Chris Gloninger, a weather and climate communication expert and former broadcast meteorologist. "But I do fear that in the case of creating towns that don't exist, that kind of damages or hurts the public trust that we need to keep building."
He noted that the wind forecast was relatively mild compared to other potential weather hazards. "If this were a forecast that would have direct impacts on public safety, or if this was a high-impact event, I don't think [generative-AI] would be used," Gloninger said.
While it's not unusual to see misinformation in our social media feeds, generative-AI errors coming from a government agency highlight the need for training and awareness as organizations across the world reckon with the pros and cons of this rapidly developing technology, said Claire Wardle, an associate professor in the department of communications at Cornell University who is focused on misinformation and AI.
"There's a lot of benefits. The problem is there's also major flaws," Wardle said. "If you're not trained to double-check for hallucinations, the location names, the logo, you're not going to realize you're making a mistake."
She noted that these kinds of errors could make the public distrust other forms of AI that are completely different from the technology you'd use to create a weather graphic.
NOAA announced in December an ambitious and advanced suite of new AI-driven global weather prediction models. Many weather experts cheered the push as a big step toward better and more accurate forecasting.
"We need to do a better job more generally in society of helping people understand the different ways that AI is being deployed," Wardle said.
Gloninger said that knowledge gap was increasingly important as AI advancements in science become more widespread.
"Just because a graphic came out wrong doesn't mean the math and science and physics behind the models that are AI generated are wrong. They're very different," he said.