AI is amplifying social media disinformation — and making Big Tech civilly liable may be the key to stemming it

• AI makes it easier than ever to spread misinformation and disinformation on social media. 
• Legal experts argue real change can only happen through new laws or the social media companies themselves.
• One expert called for amending an internet law that protects social media companies from civil liability. 

In the age of generative artificial intelligence, the public is arguably more susceptible than ever to fall victim to disinformation and misinformation on social media.

It's on social media platforms where disinformation and misinformation runs rampant. You can almost bet on it that any time major events become part of the public conversation, digital falsehoods swiftly circulate. Think of the COVID-19 pandemic and both the 2016 and 2020 United States presidential election cycles.

AI-generated so-called deepfakes are only exacerbating the problem and making it easier than ever to spread disinformation and misinformation via social media.

Legal experts told Business Insider that the only real way to combat misinformation and disinformation on social media is through the creation of new federal laws or the tech companies behind the platforms voluntarily ramping up their own self-regulation.

"AI means it's not just going to be words" that disseminate false information on social media, it's going to be videos and photos and audio recordings," said Barbara McQuade, a former US attorney and author of the book, "Attack From Within: How Disinformation Is Sabotaging America."

McQuade, a professor at the University of Michigan law school believes there needs to be new laws on the books to address this issue because "this is new technology that didn't previously exist."

"We may be reaching a point of awareness where people are beginning to understand the risk and the danger of it," McQuade said.

A recent federal assessment compiled by the US Department of Homeland Security warned of the threats AI poses to America's 2024 presidential election.

"As the 2024 election cycle progresses, generative AI tools likely provide both domestic and foreign threat actors with enhanced opportunities for interference by aggravating emergent events, disrupting election processes, or attacking election infrastructure," the analysis obtained by ABC News said.

Social media companies are protected from civil liability under a US law

Social media has largely gone unregulated since its birth nearly three decades ago. In the US, tech giants like Meta, X, and TikTok are protected from civil liability related to the content posted by their users and the companies' content moderation practices under Section 230 of the Communications Decency Act of 1996.

"It says they are not subject to legal liability, they are immune," McQuade explained. "And that probably made sense in 1996 when the goal was to foster innovation and investment, and here we are almost 30 years later, and we've seen some of the collateral consequences of this unregulated space."

So what's been the struggle for the government to address the issue of disinformation and misinformation on social media head-on? First Amendment concerns, pushback by Big Tech, and political will, have something to do with it, according to legal experts.

"It's hard to enact legislation, it's hard to define the terms" of misinformation and disinformation, it's hard to have agreement on what the proper intervention would be, and I think it's hard to craft something that's not going to have a First Amendment problem," said Gautaum Hans, a law professor, and associate director of the First Amendment Clinic at Cornell University.

"Any kind of regulation that targets speech has a very difficult barrier to constitutionality, so the issue there is you'd have to define disinformation or misinformation in a way that made it not covered by the First Amendment," said Hans.

Hans said he believes there's a "general skittishness about any legislator or any government official proposing something that could be deemed Orwellian in its attempt to try to create a regulation of protected speech."

"So most politicians, I think, are cognizant that it would be bad for their reputations to be seen as speech suppressive," he said.

Additionally, Hans noted, "there are benefits to certain political actors about the existence of misinformation."

Hans said he believes it is more likely that the remedies to misinformation and disinformation on social media will be found through the private practices of tech companies themselves instead of through the realm of law.

"I think that it is more likely to happen, and probably more effective in the long-term given the constitutional problems of legislative or regulatory intervention," he said.

Section 230 has been hotly debated for years

McQuade argued that social media companies would need to be incentivized to beef up their self-regulation on fighting misinformation and disinformation.

"I just think that you either need to put public pressure on them through consumers to make them change their behavior or through federal legislation," said McQuade.

McQuade proposed amending Section 230 in order to hold social media companies accountable under certain circumstances.

"The better course for regulating social media and online content might be to look at processes versus content because content is so tricky in terms of First Amendment protections," the former federal prosecutor said, adding, "regulating some of the processes could include things like the algorithms."

"I'm suggesting that perhaps Section 230 could be amended to provide for civil liability, you know, money damages if the social media companies didn't take certain precautions," she said.

Those precautions, McQuade said, could relate to the disclosure of algorithms and how private data is used, requiring users to label AI-generated material, and the removal of bots, which "are there to amplify false information."

"So I think that would be the way to sort of use a stick to get compliance by exposing" social media companies to legal liability for failure to comply with certain terms, said McQuade.

This would inevitably be challenged legally, and "they'd make it to the courts to see whether those laws would stick, but I think that's probably what it would require," McQuade said.

"Information is such an important resource, especially in a democracy," said McQuade. "And it seems that everyone should agree that when there is disinformation out there that is an obstacle to good government."

Section 230 has come under intense scrutiny over the years from both Republican and Democratic politicians alike.

Former President Donald Trump and other Republicans have argued the law gives Big Tech too much power to censor conservative voices, while Democrats like President Joe Biden have said it doesn't do enough to fight hate speech.

In a Wall Street Journal op-ed last year, Biden double-downed on calls to reform of Section 230.

"We need bipartisan action from Congress to hold Big Tech accountable," Biden wrote in the op-ed. "We've heard a lot of talk about creating committees. It's time to walk the walk and get something done."

Big Tech, however, scored a massive win last year when the US Supreme Court sided with Twitter and Google in lawsuits that alleged they "aided and abetted" terrorist attacks.

The decision, which was written by conservative justices, stayed out of the fight over Section 230.

Major social media companies have their own misinformation policies

Many major social media companies, including Meta, TikTok and X, have their own policies when it comes to tackling misinformation and disinformation.

For example, Meta, which owns Facebook, Instagram, and Threads, says on its website that it removes misinformation "where it is likely to directly contribute to the risk of imminent physical harm."

"We also remove content that is likely to directly contribute to interference with the functioning of political processes and certain highly deceptive manipulated media," Meta says.

Meta says it focuses "on slowing the spread of hoaxes and viral misinformation" and requires users to disclose, using its "AI-disclosure tool," whenever they post content with "photorealistic video or realistic-sounding audio that was digitally created or altered, and we may apply penalties if they fail to do so."

"We may also add a label to certain digitally created or altered content that creates a particularly high risk of misleading people on a matter of public importance," says Meta.

In a May 2024 "adversarial threat report" by Meta, the company said, "So far, we have not seen novel GenAI-driven tactics that would impede our ability to disrupt the adversarial networks behind them."

TikTok says it does not allow "harmful misinformation" on its platform and says that it has "robust policies around specific types of misinformation like medical, climate change, and election misinformation, as well as misleading AI-generated content, conspiracy theories, and public safety issues like natural disasters."

X, the social media site formerly known as Twitter, says on its website that users "may not share synthetic, manipulated, or out-of-context media that may deceive or confuse people and lead to harm ('misleading media'). "

"In addition, we may label posts containing misleading media to help people understand their authenticity and to provide additional context," X says.

These policies, though, are not always enough.

Earlier this year, graphic AI-generated images of Taylor Swift went viral on X, with a post garnering 45 million views before it was finally taken down 17 hours later.

X, which is owned by Elon Musk, blocked searches for the popstar in the aftermath.