Connecticut Agency AI Image Fail Goes Viral

An AI Image Gone Wrong

It was a social media post that quickly drew the wrong kind of attention. The Connecticut Department of Energy and Environmental Protection (DEEP) recently shared an image on Facebook that was meant to promote hunting safety. However, eagle-eyed internet users immediately spotted the tell-tale signs of an AI-generated image. The hunter depicted had an extra finger, his bow was missing an arrow, the text on his hat and vest was nonsensical gibberish, and even the turkey he was targeting appeared to be wearing its own tiny safety vest.

This image, created with artificial intelligence, was posted online by the Department of Energy and Environmental Protection. Credit: CT DEEP

The peculiar image was quickly shared, with users on Reddit and Facebook pointing out the numerous flaws.

The Agency's Response and Public Backlash

In response to the online feedback, DEEP took the post down on Monday and replaced it with what they called a "better and less confusing" image. The agency admitted that they used Canva AI to create the graphic and that a mistake was made.

"We did not apply our usual level of scrutiny to the image used in this particular post before posting it," DEEP said in a statement. They clarified that using AI-generated images is not their standard practice, as they typically use real photos from the field or stock images. The original post also failed to disclose that the image was created by AI.

The public's criticism went beyond the quality of the image. Some users pointed to the known environmental impact of training AI models, while others raised concerns about a government agency using AI instead of hiring a local photographer.

AI in Government: A Look at State Policy

The incident brings Connecticut's own policies on artificial intelligence into the spotlight. The state has an AI Responsible Use Framework that requires state agencies and their vendors to ensure AI produces "accurate and verifiable information." While the framework mentions the use of AI for generating visual content, it doesn't offer specific guidance on AI images.

Interestingly, the state is required to publish an annual inventory of the AI tools it uses. Canva, the platform DEEP used, was not included in Connecticut's 2024 list. Following the incident, DEEP has stated that any "future uses of AI in the creation of images will be noted with the post."

The Underlying Message and Final Thoughts

While the post was described as "a little bit embarrassing" by Chas Catania, the Treasurer of Friends of Connecticut Sportsmen, he stressed that the core message shouldn't be lost. The original post was intended to inform the public about the rules for the fall hunting season, including the crucial safety measure of wearing fluorescent orange.

"I think what’s very important is that we don’t lose the underlying message, and that is, you know, they’re encouraging safety," Catania said. The incident serves as a public lesson for government agencies navigating the new world of AI tools, highlighting the need for careful scrutiny and transparency.