Vice President Kamala Harris has long been the target of racist and misogynistic attacks online as a woman in politics who is also Black and South Asian. As she seeks the Democratic nomination for president, the gendered, sexualized and racial disinformation against her could get worse — and experts warn there may be less comprehensive oversight by tech companies to stop its spread this election cycle.
In the days since President Joe Biden announced he would not seek reelection and endorsed Harris to be his successor, right-wing media personalities and others have posted disinformation and misinformation — intentional and unintentional spreading of false information, respectively — about Harris in memes, videos and writings online.
The posts, shared on popular social media platforms, have primarily focused on sexualized and racist narratives about Harris, including claims about her dating history and her eligibility to be president. Disinformation experts who spoke with The 19th said highlighting or linking to more specific disinformation about Harris risks amplifying it.
The 19th thanks our sponsors. Become one.
Nina Jankowicz, cofounder and CEO of the American Sunlight Project, an organization that tracks how deceptive information is spread and its threats to democracy, believes the tech companies that run these popular platforms are not doing enough to stop the spread of disinformation about women candidates like Harris. She noted that some social media moderation systems can miss detecting posts being shared right now about Harris in part because the problematic messages are often coded to hide their meaning.
“People have free speech. People can be critical of people in politics, whether they’re men or women. But what we’re talking about here are posts that go against the platforms’ own terms of service — in terms of violent rhetoric, in terms of sexualized rhetoric, in terms of gender-based or ethnicity-based or race-based harassment,” she said.
Tech companies that run social media platforms have made some public commitments to address disinformation generated by artificial intelligence in election-related content. They’ve also faced scrutiny on the scope of their efforts on overall content moderation.
On Instagram, which is owned by Meta, at least one account shared several memes about Harris that implied a sexual narrative about her. The company, which has a plan for online manipulations around elections, later removed the content. Manipulated audio of Harris also quickly circulated on TikTok at the start of the week and was shared hundreds of thousands of times before TikTok removed the audio, which falsely shows Harris slurring remarks. A representative for the platform told the progressive research group Media Matters for America that it is “actively and aggressively removing this content.”
Jankowicz pointed out that many of the altered images and memes being shared about Harris are not AI generated, but instead “cheap fakes” — in other words, content that is created with easy-to-access software or other simple means that also makes it more difficult to detect sometimes. She worries that the sustained online attacks against Harris and other elected women will lead some who are considering politics to give up on running for office in the future — a difficult-to-track ripple effect that could ensure politics remain heavily male and White.
“Leaving aside the falsity of these narratives, even that is damaging,” she said. “It’s damaging to women’s political participation writ large.”
Melanie Smith, director of research for the Institute for Strategic Dialogue, an organization that analyzes polarization, extremism and disinformation worldwide, has done extensive research on gendered online attacks against women in politics. Some of their research findings show that women of color are disproportionately impacted by abusive content online.
Smith said the landscape of online moderation by tech companies has changed for the worse even since the 2020 presidential election and the 2022 midterms. Some tools that were available then to track disinformation will not be available in the fall. Communication between researchers and the major social media platforms about data has also diminished, according to Smith, who said companies don’t share as much information as they used to — leaving researchers to do a lot of the tracking manually. That and less available information have made it harder for them to find and expose complete trends.
Citing the recent example of manipulated audio of Harris, Smith said the use of artificial intelligence on social media platforms remains a concern.
“We, right now, do not have the kind of counter infrastructure in place to detect every instance of a deep fake video or a manipulated audio,” she said.
Jankowicz, herself the target of disinformation, is the coauthor of a study conducted in 2020 and released in early 2021 that analyzed gendered abuse of women candidates. The analysis looked at 13 women across six social media platforms and more than 336,000 pieces of abusive content shared by over 190,000 users over a two-month period. Harris accounted for 78 percent of the total recorded instances of gendered abuse and disinformation.
“Unfortunately, Harris has been targeted with these as long as she’s been in politics. The internet makes them more potent,” added Jankowicz, who wrote “How to Be a Woman Online: Surviving Abuse and Harassment, and How to Fight Back,” published in 2022. “These narratives exist and they are being used in a coordinated way in order to undermine the idea that she is fit for office.”
Smith said there’s power in the general public being aware of the intention here.
“They’re not trolling her just to troll her. There’s a broader effort at play that is connected to undermining women’s participation in democracy and women holding office,” she said. “There is this bigger agenda.”
Former President Donald Trump claimed ahead of his party’s convention this month that he would try to share a message of unity to potential voters, though his words and policy proposals have been rife with animosity toward immigrants and transgender people. He has also repeatedly criticized Harris — and not just over differences in policy positions. In early July, he posted on his Truth Social account a veiled reference to Harris’ dating history on a social media post as a means of tying it to her professional ascension.
On that same platform, Trump this week called the vice president — who is also a former U.S. senator, attorney general and district attorney — “dumb as a rock.” Trump has also tried to make fun of Harris’ laugh, which his campaign included in a memo released about her record in public office. The laughter also closes the first attack ad by a pro-Trump political action committee, released this week.
“That focus on style over substance is an attempt to diminish beliefs that she is ‘presidential,’” Kelly Dittmar, an associate professor of political science at Rutgers University and the director of research at the Center for American Women and Politics (CAWP), said during a call Tuesday with reporters. “All of these attacks and the degree to which they resonate reflect the persistent dominance of masculinity and Whiteness in presidential politics, and we see candidates and campaigns double down on these status quo dynamics in messaging and presentation.”
Some politicians used racist tropes to claim Harris was unqualified to run for president. Rep. Tim Burchett, a Republican from Tennessee, said on Monday on social media and during a CNN interview that Harris was tapped as vice president through DEI — which stands for “diversity, equity and inclusion” but has been coopted by detractors as a racist message about a person’s qualifications for a job. Rep. Marjorie Taylor Greene of Georgia made a similar claim on Tuesday on social media. Politico reported that House Speaker Mike Johnson has instructed Republicans to stick to messaging about Harris’ policies instead of her gender and race.
It’s not just mainstream online platforms where racist and sexist messaging is running rampant. In the first 48 hours after Harris announced her intention to run for arguably the most powerful job in the world, hate speech about her rose sharply on websites where users are known to share right-wing conspiracies, according to an analysis by the Global Project Against Hate and Extremism, an organization that has been tracking how political discourse is shaping the potential for political violence.
“Women politicians, particularly women of color, have long experienced harassment, misogyny, and racism, sometimes to the point of being driven out of politics,” Wendy Via, a cofounder of the project, said in a statement to The 19th. “What is shameful is that this kind of vile rhetoric in our political discourse is completely unsurprising at this point. We can’t allow ourselves to become numb to the dehumanization of women, especially as we’re being used as pawns in a political power grab and the undermining of democracy.”
Hillary Clinton, the first woman to win a party’s nomination for president, talked about her own experiences in facing “sexism and double standards” in an opinion piece Tuesday, where she also spoke of the attacks she endured when she ran for president and the realities that Harris will face in the weeks ahead: “She and the campaign will have to cut through the noise, and all of us as voters must be thoughtful about what we read, believe and share.”
The potential impact of gendered disinformation includes political violence, said Joanna Lydgate, cofounder and CEO of States United Democracy Center, a nonpartisan organization that advocates for advancing free, fair and secure elections. She said hateful rhetoric increases political polarization and the tolerance for political violence — with real-world impacts on elected officials like an assassination attempt, a kidnap plot and doxxing. She emphasized the group’s work in training law enforcement to be responsive to these threats.
“Whether it’s voter intimidation, election official intimidation, threats, harassment — this is something law enforcement leaders take really seriously,” Lydgate said. “It is increasingly becoming part of their job. And there are absolutely consequences for people who step outside the bounds in this way.”
Smith, of the Institute for Strategic Dialogue, noted that representatives for social media platforms have indicated in public testimony that they use automated moderation for some posts that they suspect violates their terms of agreement. She worried, however, that the strategy won’t capture some of the messaging that can be coded to avoid detection.
“I’ve been doing digital research for a good while. I still don’t believe that machines can really understand the context of things like hateful rhetoric,” she said. “They can help, but you ultimately need a human to decide whether something is violating a policy or not, and contextualizing that content is really important.”
Smith said she expects the memes and posts about Harris to worsen as a more organized disinformation operation takes shape. Harris, who has secured the support of enough delegates to clinch the Democratic nomination for president, would make history as the first woman president and the first in the role who is a woman of color.
“We are likely to see a period of throwing spaghetti at the wall in terms of testing different conspiracy theories about her, different disinformation narratives,” Smith predicted. “And the people who have a vested interest in undermining her will seek to see what’s sticking with their audiences.”
Lydgate said it’s important to note that many women who face gendered attacks remain committed to serving in public office. She called on political leaders to speak up.
“Our political leaders have a real choice at this moment,” she said. “They can continue to raise the temperature or take it down. There is such a responsibility but also real power right now for them to stop the cycle of political violence that we’re seeing in our country.”
