In 2021, Chanel Contos was one of the brave Australian women speaking up about a particularly dark subculture of harassment and abuse in Australia.
Contos has made consent the centerpiece of her campaign. Ironically, many of the technology platforms used in her campaign only give lip-service to the concept of consent.
Contos has compiled and anonymized thousands of testimonials from young female victims. According to reports, Contos and helpers have manually copied these testimonials from different messaging tools, removing any identifying information in the process. This type of content moderation is extraordinarily difficult. Contractors who do content moderation for social media platforms have made widespread complaints about the psychological impact. I recently wrote a blog about the Molly Russell inquest, where the expert psychiatrist confessed that he had problems sleeping after looking at the content that pushed Molly Russell over the edge. Reading about the Debian Day suicide and the Prince Alfred research both have a similar impact on people, both the researchers and the people following the blog posts.
During the campaign, many media web sites reported on Contos' connection to UCL. Yet I could find no official statement from Contos or UCL claiming that this was an official research project. If the project was officially operated by UCL then it would require approval from the UCL ethics board and it would require extensive controls for human subject research. When high school students participate in human subject research, there are even stronger controls and parental consent requirements. Contos' work is in a gray area. She did not claim it was a UCL project, nonetheless, given all the media emphasis on her status at UCL, many participants may have given it the same level of trust as an authorized research project.
Furthermore, if identifying data is hacked or leaked, this may deter women from participating in research projects in future. Women's health is an area that is already neglected in many ways and it would be disappointing if a well-meaning campaign created barriers for future researchers.
Contos' Wikipedia page talks about interaction with the police service. This is also a gray area. It would appear to be quite logical and sensible to give every participant a list of all relevant support services. On the other hand, if participants feel that research is too close to the police, they may be less willing to provide sensitive information.
One of the startling revelations is that the early version of the consent web site included the names of two other developers, one of them connected to Palantir and Google. Palantir is an official subsidiary of the CIA specializing in big data.
It is well known that in the hunt for Osama bin Laden, the US used a fake vaccination campaign to locate bin Laden. Tactics like this make it harder to convince the public about genuine vaccines. The downside of those tactics only become obvious in recent times with the emergence of anti-vax protests against the Covid vaccine. The type of incident documented in Miss Contos' petition is so widespread that I don't think they need help from the CIA to make them up.
While the campaign appears to be entirely genuine, the social media companies will have gathered extensive meta-data about the participants and the people they are accusing. A lot of this data is automatically harvested by intelligence agencies and may be used to blackmail people in the future.
The police may feel an obligation to mine the same data. Eighty percent of perpetrators have probably only been subject to one complaint. Twenty percent are probably committing a crime every month, maybe even every weekend. Police may feel obliged to mine the meta-data to try and identify repeat offenders.
In the case of the Boston College tapes, the institution fought tooth-and-nail to keep the tapes away from the police but eventually they were forced to hand them over. For an unofficial research project conducted over social media, it would be far easier for authorities to obtain the meta-data, if they haven't already done so.