Misinformation roiled the 2016 and 2020 elections. Here’s what to expect as 2024 races heat up.

“On the one hand, it feels the same,” said Claire Wardle, cofounder of the Information Futures Lab at Brown University’s School of Public Health in Rhode Island, of this election. “But actually, I think the world looks quite different four years later.”

Here is a sample of what misinformation researchers are looking out for this presidential election year.

Like the Biden deepfake robocall, other political misinformation tactics this year will likely try to push voters to not vote rather than toward a certain candidate, experts said.

“If I was going to try and do something to influence this election, whether I was domestic or foreign, I am not going to do anything really that tries to persuade people in terms of their voting. I’m going to stop people from voting,” Wardle said.

The sources of these efforts could come from China and Russia, which have been documented as originators of election misinformation in the past, said Matthew Baum, a professor of global communications at Harvard University. But he added that there is plenty of domestic misinformation as well.

“These sorts of campaigns are more about dispiriting the other team,” he said. “More than anything, get the other team’s supporters to not show up.”

Generative AI lowers the barriers to entry for carrying out misinformation campaigns, researchers said.

“I don’t think that AI is necessarily going to make misinformation better, in the sense of making it more persuasive,” said Adam Berinsky, a political science professor at MIT. “But it’s easier to create misinformation.”

For example, the man who took responsibility for the “deepfake” Biden robocall in New Hampshire told NBC News creating the recording took less than 20 minutes and cost $1.

“If this is the kind of thing we’re going to see from a no-hope, no-chance campaign in New Hampshire, think what we’re going to see when there’s actual stakes,” said Laura Edelson, an assistant professor at Northeastern University who studies misinformation.

AI-generated images, audio, and video can be used to try to influence anything from a presidential race to local contests.

Wardle said that creating deepfake audio of local personalities like faith leaders or high school coaches could be “very persuasive at a local level” and also hard to monitor, which could make such efforts successful.

Generative AI has spurred a lot of hype, but researchers said they’re also still expecting a more old-fashioned type of misinformation: candidates’ false claims.

“I’m still a bit more concerned about false narratives that are coming from places like our political elites,” said Katherine Ognyanova, an associate professor of communication at Rutgers University in New Jersey.

Researchers said misinformation comes from both sides of the political spectrum but disproportionately from the right. That includes Trump, who said without evidence that the election was rigged in 2016 and then repeated falsehoods of vote fraud when he lost to Biden in 2020. His well-worn slogan of “stop the steal” morphed into failed efforts to overturn the election, which researchers said people should remember this cycle as a symbol of the potential danger of spreading lies.

“I think we are being exceptionally naive if we don’t think about what will happen the day after the election here, irrespective of the result,” Wardle said.

A Trump campaign spokesperson did not respond to a request for comment.

Young people’s media diets make them especially prone to encountering misinformation. Social media can serve up more falsehoods than traditional outlets.

But anyone, regardless of age, can be misled, Berinsky said.

There are lots of fears about TikTok, which shows short videos and is the social media app of choice for Gen Z. One-third of US adults under 30 get news from TikTok, according to the Pew Research Center.

Some researchers, such as Baum, said TikTok could become a vector for misinformation. But others, such as Boston University assistant professor of journalism Joan Donovan, said TikTok has stronger controls to prevent misleading and hateful content.

“Using TikTok to spread disinformation is much harder than it is for using social networking to do that work,” Donovan said. “There’s very strong moderation on TikTok around certain political themes and hate speech.”

Meanwhile, Meta, the parent company of Facebook and Instagram, has announced plans to de-emphasize news and political content over the past year. And X, formerly Twitter, under owner Elon Musk, has gutted its trust and safety team. That leaves researchers worried about the content people could see on social media.

Trump and far-right groups have, and will continue, to spew false and anti-immigrant narratives to try and court Latino voters, researchers said. That comes as the Latino population has continued to grow.

Many Latino citizens are also swing voters. Some misinformation in swing states has specifically messaged toward them. That includes in 2020, when Spanish-language ads on Facebook falsely labeled Biden a communist.

But “it’s not the case that Latinos are more susceptible” to misinformation than other demographic groups, said Roberta Braga, executive director of the Digital Democracy Institute of the Americas in Washington, D.C. The volume of misinformation increases during election years and affects everyone, including Latino voters, she said.

Braga said she expects “similar but more ingrained, false and misleading narratives” this cycle, but also believes Latinos have shown more resistance to misinformation because people are more aware of its patterns.

Aidan Ryan can be reached at aidan.ryan@globe.com. Follow him @aidanfitzryan.