CHARLOTTESVILLE, Va. (CBS19 NEWS) -- Be careful believing what you see online. Political experts at the University of Virginia are warning that campaigns could make attack ads using what's called "deepfake" videos and recordings.

The UVA Center for Politics says that while artificial intelligence could benefit other parts of our lives, these computer-generated videos could have negative effects on democracy.

"We have political consultants, campaigns, candidates, and even members of the general public that are just forging ahead with the technology without fully understanding how it works," said Carah Ong Whaley with the UVA Center for Politics.

According to Ong Whaley, deepfakes are computer-created manipulations of a person's voice or image that use machine learning to create content that appears as if it were real. She says this can be used for good, but also for bad.

"We just don't have a sense yet of the potential impacts and the way it could magnify existing biases or existing harms," she said.

She says elections can become a victim of deepfakes.

"That could really just target specific demographics and either suppress votes or spread disinformation at a scale we have never seen before," Ong Whaley said.

Ong Whaley says politicians like former President Donald Trump and Republican presidential candidate Florida Governor Ron DeSantis have already made use of them. But they aren't the only ones.

"We've seen the use of doctored videos and images in the past. I think what is new is now anyone could do it, anyone could put it out. There are no guardrails in place to prevent its use and there are no disclosure requirements," she said.

She says Congress needs to address the issue.

"Congress has been working on legislation. But nothing is in place yet. We need Congress to act. We need a regulatory regime and we really need to have something in place yesterday," Ong Whaley said.

Even if Congress does do this, it wouldn't have any effect on potential foreign actors.

"It doesn't apply to non-state actors, it doesn't apply to foreign actors so the potential for harm is still there. Generative AI poses severe dangers and challenges not just for the US, but for democracies around the world," said Ong Whaley.

The Federal Election Commission has opened public comment for people to express their concerns about deepfakes. This is open through mid-October. To submit feedback, click here and search for REG 2023-02. Comments need to be submitted by Oct. 16.