The risk of AI-generated "deepfake" content messing with the 2020 election has prompted two US senators to call on Facebook, Twitter, and YouTube to come up with industry standards to detect and stamp out the potential threat.

On Wednesday, US Senators Marco Rubio (R-Florida) and Mark Warner (D-Virginia) sent letters to major social media platforms—including Reddit, Snapchat, TikTok, Twitch, and LinkedIn—about the dangers of deepfakes spreading misinformation to the public.

"We believe it is vital that your organization have plans in place to address the attempted use of these technologies," the senators wrote. "We also urge you to develop industry standards for sharing, removing, archiving, and confronting the sharing of synthetic content."

Although anyone can edit a photo, today's AI-powered programs can doctor images and videos to a new, realistic extreme. Capabilities include swapping your face on to a famous actor's body, taking existing footage and manipulating the subject's face to say something else, and generating lifelike headshots of people who don't actually exist.

Deepfake technologies can also be applied to audio. Recently, an AI-based software program was used to mimic a company CEO's voice to trick an employee into making a fraudulent wire transfer.

Sens. Rubio and Warner are worried the same technologies will be used to pump out misinformation that'll dupe the public and quickly go viral on social media with nothing to stop it. "If the public can no longer trust recorded events or images, it will have a corrosive impact on our democracy," the senators wrote.

The major tech companies agree about the potential danger posed by deepfakes, but Rubio and Warner claim there's been "limited progress" in creating industry-wide standards to tackle the problem, and have asked the social media platforms to answer seven questions regarding their policies to stop deepfakes and warn users about them.

Last month, Facebook and Microsoft kicked off a contest to recruit researchers who want to develop tools to detect AI-generated videos.

Google, which owns YouTube, has also released a dataset of 3,000 deepfake videos it's created to academics in the hopes they can create systems to detect fabricated imagery from the real. "We're committed to investing in the technology and teams that enable us to understand this challenge and maintain the integrity of our platform," YouTube told PCMag in response to the senators' letter.

Whether the companies will succeed at regulating the content is another matter. Facebook, YouTube, and Twitter have all routinely faced criticism for failing to stop misinformation and hate speech while also incurring the wrath of free speech advocates who've blasted them for online censorship. Facebook CEO Mark Zuckerberg previously said his company is developing a policy around deepfakes. But the social network has largely refrained from deleting fake news and instead focused on demoting the content, and placing it next to fact-checked news articles.

Rubio and Warner say they advocate for a "clear strategy and policy in place for authenticating media." Deepfake content should also be labeled and archived so researchers can track disinformation campaigns, especially if a foreign government may be behind the propaganda effort.

"The threat of deepfakes is real, and only by dealing with it transparently can we hop to retain the public's trust in the platforms it uses, and limit the widespread damage, disruption, and confusion that even one successful deepfake can have," they wrote.