AI-generated news offers unchecked path to misinformation as elections approach

What to trust and what not to trust, that is the question - so go with your gut feeling and vote as you want, not how others think you should - vote RFK!!

AI-generated news offers unchecked path to misinformation as elections approach

John Cardinale

National Investigative Correspondent

A new issue is emerging as unchecked artificial intelligence continues to permeate online: There's an explosion of websites that look like news sites — but they're automated through AI.

Hearst Television's National Investigative Unit spoke with several secretaries of state, the people ultimately responsible for conducting elections in their states, who said there are real concerns that stealth uses of AI in this manner could influence the outcome of elections across the country.

Maine Secretary of State targeted

Maine Secretary of State Shenna Bellows knows about the impact of misinformation all too well.

According to online articles from what appeared to be a news website, she was arrested, served time at Guantanamo Bay — and was executed. But none of it was true.

"We were in this awkward position of having to explain, to some media outlets, like USA Today, also to some voters, 'No, I'm alive and well. I'm, you know, here in Maine doing my work,'" Bellows said.

Hearst Television

Maine Secretary of State Shenna Bellows

Those stories clearly can have a lot of power.

As the secretary of state, Bellows is the one in charge of running Maine's elections.

That is why she has been targeted, both by people and AI, with articles about her filled with incorrect information.

Stories on these websites are written by bots, allowing for potential misinformation and disinformation to be published with little to no human oversight.

A growing problem

AI-powered websites are becoming more of a problem, according to Steven Brill, co-founder and CEO of NewsGuard, an organization that provides tools to counter misinformation.

"We started with 40 or 50, and that has just exploded," Brill told the National Investigative Unit. "A month from now, it could be 1,200. If I'm talking to you right before Election Day, I'll bet it's over 2,000."

Brill's team is tracking more than 1,000 websites run by AI that spread false narratives. The websites the organization is tracking appear to be predominantly automated, using a script to scrape authentic news sites and rewrite the articles using AI, twisting the facts.

These sites can gain significant traction on the internet, especially through social media.

"They can get thousands, hundreds of thousands, millions of views," Brill said. "So the website itself is really just kind of a marker of false legitimacy."

Many of the websites are targeting the election, according to Brill.

These websites can also be difficult to distinguish from legitimate news sites because they sometimes have generic names that are close to trusted news brands and are laid out in a way that mimics them. Some of these AI-powered sites even have multiple sections, articles, author bylines and advertisements.

"You can only imagine how that is going to accelerate as we get closer to the election," Brill said.

What's being done to stop AI and misinformation online?

Social media websites like Facebook and X, formerly known as Twitter, have vowed to flag articles that contain misinformation and disinformation.

Some states are even passing laws, adopting resolutions and/or enacting legislation to address the use of AI in different ways.

Explore the map below to discover the laws across the nation.

How can you tell if a website is AI-generated?

AI is constantly evolving, so as time goes on, it will be harder to distinguish these sites from human-written ones. That's why it's important to choose trusted news sources.

Below are basic tips and red flags when evaluating news sites.

Check the "About" and "Privacy Policy" pages

Trusted news sites almost always have "About Us" and/or "Privacy Policy" pages, which offer transparent information for readers about the organization and its ethics. Readers can find ours here.

But some AI sites have overlooked key details on these pages. Some sites may say these pages are still being developed. For example, "This website was founded in [date] by [Your Name]." The more generic the language, the better the chance that there is some automation involved. It's rare, but some might actually state the website is generated by AI or that the content has been written for "satire" and isn’t based on facts.

No background on the author?

Take the author's name and type it into Google. If they don't have any past work, that's a red flag. Experts say AI "authors" will produce more content as time goes on, so be critical when looking at past work.

If the article attributes the writing to "Admin" and "Editor," or has no bylines at all, that's another red flag.

Readers may also want to examine the author more closely. The Hearst Television National Investigative Unit found at least one of these websites using the author bylines of real reporters, but further investigation showed no real connection between those reporters and the suspect articles.

Scan for text that seems too formal or out of place

"In c0nclusion," is a common phrase that appears at the end of many AI articles, but it's something human journalists rarely, if ever, use.

Experts say artificial intelligence systems are meant to produce "helpful" responses, and that sometimes includes language that isn’t conversational. If the language used feels too formal or too jargony, it's another red flag.

Sentences that are repeated multiple times

AI-generated articles sometimes contained repeated concepts or even whole sentences in their copy. Human journalists and editors would likely have removed or refined these statements.

Check the context

AI lacks a human understanding of the world. So, if the story seems unable to grasp a larger context or is missing the point entirely, think twice about it.

Experts say that AI tries to predict the next word in a phrase or sentence, not produce facts. These articles often miss out on broader context.

Check the sources

A legitimate news article may frequently use citations and hyperlinks to link you to sources of information.

Also, if the headline is not posted anywhere else, be skeptical. Look for other sources, particularly ones that you've heard of and trust.

WISN

Me: Mind you, I won't be checking to see if a post is bot related - if I like it I'll read it and I might pass it on, if it does not conveigh the idea I want to pass on, then I won't.