As a rape investigation at Cornell University spiraled into a national firestorm, some people went online to seek the identity of the alleged victim.
Social media posts on X that attempted to reveal the name of the woman at the center of the case, identified in court filings only as Jane Doe, were viewed more than 87 million times over two weeks, according to data reviewed by The Washington Post.
Facebook, Instagram and TikTok attempted to limit the spread of content purporting to identify Jane Doe. All three platforms blocked searches for an alleged name for Jane Doe, although they appeared to allow some posts containing that name to stay online.
Things were different on X, the platform owned by Elon Musk. The service allowed searches for the alleged name for Jane Doe and the company’s chatbot, Grok, also posted it in replies to users at least 300 times, according to data compiled by researchers at the Center for Countering Digital Harm.
X posts claiming to name or identify Jane Doe received 87 million views in a two-week period after local prosecutors said on Sept. 28 that they would reopen the investigation, according to the researchers.
“A responsible organization would start thinking … does this break our rules? If not, something’s wrong with our rules and we need to fix them,” said Imran Ahmed, the chief executive of the nonprofit, which was previously known as the Center for Countering Digital Hate.
“The problem is that we have accepted a lazy untruth, which is that there is nothing we can do about these platforms, and we have no say in the way they behave,” he said.
X did not respond to a request for comment. The platform’s privacy policy prohibits sharing some types of private information without consent, such as a person’s home address. It lists others, including names and birthdates, as data that “we don’t consider to be private.”
Since Jane Doe filed a lawsuit alleging that she was drugged and raped at a fraternity house party in 2024, she has been harassed by anonymous callers and subjected to doxing, where people attempt to reveal personal information online, according to her lawyer, Thomas Giuffra.
He has also said that police appeared at her parents’ home with guns drawn after a false report of an ongoing murder, a tactic known as swatting.
Doe’s suit seeks damages from seven fraternity members and Cornell University, among others. Several young men who share names with those accused in the lawsuit have also been targeted by harassment. Giuffra has said that online users have claimed to identify at least three women as Jane Doe on social media.
The online response has been so ferocious that the judge in the case ordered several key documents, including the initial complaint, to be sealed over concerns about doxing.
On X, the platform’s artificial intelligence chatbot Grok contributed to the flood of posts containing alleged information about Jane Doe. Grok, which can publicly reply to users of the platform, appeared to pick up the alleged name from the online chatter.
In some cases Grok provided the purported name in response to questions about Jane Doe. In others, the chatbot declined to confirm user requests to verify Jane Doe’s identity and noted that alleged information for her was from unverified online speculation.
When some users asked Grok to confirm their speculation about a partially redacted version of Jane Doe’s signature in court documents, the chatbot obliged. “Highly probable,” Grok said of one guess. “High plausibility,” it responded to another.
Grok’s public posts in response to X users have previously caused controversy and forced a reaction from the platform. After users prompted the chatbot to digitally edit photos of women and girls into sexualized content, X said it had limited the feature.
The first traces of X posts attempting to identify Jane Doe appeared on Sept. 29, the nonprofit’s researchers said — the day after New York state prosecutors announced that they would reopen their investigation into the alleged rape.
The posts peaked around the weekend of Oct. 3. Over the course of 10 days, more than 2,700 distinct users posted, nearly 90 percent of them with five posts or fewer.
Meta, which owns Facebook and Instagram, said its policies do not allow content that reveals the identity of people in legal cases where that has been restricted from public disclosure. TikTok did not have a comment.
Limiting the search of a specific phrase — as TikTok, Instagram and Facebook did for an alleged name of Jane Doe — doesn’t prevent potentially harmful content circulating. Users still get the vast majority of their content from their algorithmic feeds, said Laura Edelson, a professor at Northeastern University.
A search block creates an “endless cat-and-mouse game,” said Katherine Cross, a PhD candidate and lecturer at the University of Washington Information School. Users could easily get around the restrictions by substituting other characters in for letters, or swapping the order of keywords, she said.
Platforms should also take down problematic posts, Cross said in an email.
“The harm is already being done, and there is something profoundly inadequate and tokenistic about the search block,” she said. “It is, doubtlessly, dissuading some people but it’s a Band-Aid on a sucking wound.”
The post The online campaign to unmask a Cornell rape accuser reached millions appeared first on Washington Post.
