advertisement
The story of an influential BJP figure having exclusive privileges in regards to which Instagram posts are taken down has, over the last ten days, embroiled an Indian news organisation and a tech giant in a controversy that’s one for the books (or OTT).
In an exclusive report on 10 October, The Wire alleged that Amit Malviya, the head of the ruling party’s IT Cell, had reported certain posts by a satirical Instagram account and they were all taken down solely at his discretion. Why? Because that was one of the perks of being classified as an XCheck user like Malviya, as per the report.
To support its claims, The Wire produced screenshots of a purported internal report that had been furnished by a source which The Wire claimed it had within Meta. Two days later, Meta denied the allegations in a statement on its official website.
While the reports may be out of sight, the unprecedented and messy affair is hardly out of mind. What are the alleged discrepancies? Who’s responsible for the alleged fabrications? What is up with Meta’s content moderation? Will anyone come clean? Here’s what we know for sure and all that remains to be clarified.
A total of seven Instagram stories and posts that were put up by the Instagram account @cringearchivist were reportedly taken down over a span of seven months. The most recent of the removed posts, dated 19 September, focused on a man worshipping the idol of Uttar Pradesh Chief Minister Yogi Adityanath.
It’s no secret that Meta’s AI-driven content moderation system often gets things wrong. Several users consistently wonder why certain posts are swiftly taken down by the platform for reasons that don’t add up. To make matters worse, Meta isn’t fully upfront on how it uses its automated tools.
While data pertaining to Meta's automated or "proactive" moderation is disclosed in periodic compliance reports, it does not provide details on how much of the content that was flagged proactively did not, in fact, violate the platform's Community Guidelines upon further review.
Meta hasn’t disputed that the XCheck or cross-check programme exists. “Our cross-check system was built to prevent potential over-enforcement mistakes and to double-check cases where a decision could require more understanding or there could be a higher risk for a mistake,” the company said in its statement.
Notably, as per Meta, XCheck is not limited to only celebrities or politicians and could also include accounts of prominent journalists as well as human rights organisations.
The Wall Street Journal’s 2021 investigation, that was also based on leaked internal files, had revealed that XCheck users enjoyed blanket exemptions from the platforms’ content moderation rules.
A month after the XCheck investigation by WSJ broke, the Facebook Oversight Board took the company to task and demanded greater transparency, specifically on XCheck. “In the Board’s view, the team within Facebook tasked with providing information has not been fully forthcoming on cross-check. On some occasions, Facebook failed to provide relevant information to the Board, while in other instances, the information it did provide was incomplete,” it said.
Two Instagram posts, including the story featuring Yogi Adityanath, were reportedly taken down for violating the platform's guidelines on content related to nudity and sexual activity. But the clip shows a man performing aarti of an idol of Yogi Adityanath.
A post/story can be taken down on these grounds, as per Instagram's guidelines, if it includes "photos, videos and some digitally-created content that show sexual intercourse, genitals and close-ups of fully-nude buttocks. It also includes some photos of female nipples."
However, in the Instagram story related to Yogi Adityanath, there was neither any visible nudity nor anything showing explicit or implied sexual intercourse.
On 18 October, Instagram user @cringearchivist claimed that the story related to Yogi Adityanath was quietly reinstated. This was on the same day that The Wire suspended its reports.
Are both meta.com and fb.com operational? Calling Meta spokesperson Andy Stone’s email in The Wire’s 11 October report "a fake," Meta’s chief information security officer (CISO) Guy Rosen tweeted, “The supposed email address from which it was sent isn’t even Stone’s current email address, and the "to" address isn't one we use here either.”
However, an Intercept journalist named Sam Biddle said that he had received an email from Stone on 31 August with ‘fb.com’ as the email domain name. Former Facebook CISO Alex Stamos said, “Clearly, the company's O365/ProofPoint instance is setup to receive emails for either domain and various employees are sending from different domains.”
Twitter user @cringearchivist said that the Instagram account @cringearchivist had gone private in April. The Twitter account posted screenshots purportedly showing the Instagram account as having been set to 'private' 24 weeks ago or “since the last week of April.”
The admin(s) of the Twitter account, who claim that they also run the Instagram account with the same username, said that Amit Malviya’s official handle did not follow the Instagram account @cringearchivist.
If the Instagram account @cringearchivist was a private account since the last week of April and if Amit Malviya’s official account wasn’t one of its followers, could Amit Malviya possibly have reported the post? Was it brought to his attention by one of @cringearchivist's followers?
The Wire's reports did not address this and therefore, this is something that is still an unknown.
Generally, one of the ways in which a user in India can report an Instagram post/story of a private account, even if they do not follow the said account, is by submitting a complaint to Meta's India Grievance Officer through an online contact form or sending it to the platforms' common physical mail address.
Given the immediacy with which @cringearchivist’s posts were reviewed and removed, was this actually Meta’s automated moderation system at work?
But this supposed clarification on why the posts were taken down only leads to more questions. Did the process involve both automated tools and in-house human reviewers? To what extent is human oversight exercised over Meta's automated moderation?
When exactly was the post pertaining to Yogi Adityanath purportedly taken down? This question was posed by Pranesh Prakash, the co-founder of digital research organisation Centre for Internet Society, to the Twitter account @cringearchivist.
The account replied that the post was taken down between 2:15 pm and 3:22 pm. The millisecond timestamps in the purported internal report accessed by The Wire were between 5:12 pm IST and 5:15 pm IST on 19 September 2022.
This might qualify as a potential discrepancy but @cringearchivist refused to specify the timezone, allegedly fearing that it would compromise The Wire’s sources. Hence, it was not possible to corroborate the timestamps in The Wire’s report, at least not in this manner.
In his tweet thread refuting the allegations by The Wire, Rosen said that the internal report appears to be a fabrication, questioned the URL in the report, and said, “The naming convention is one we don't use.” Facebook whistleblower Sophie Zhang raised a similar concern that the internal report “seems too formatted/formal (unless a tool was made.)” So, what does an internal post-incident review report look like?
Even as The Wire's reports remain suspended, the laundry list of unanswered questions makes us wonder if we'll ever know the whole story.
(Update, 23 October: This report was updated with The Wire's statement retracting its coverage.)
(At The Quint, we question everything. Play an active role in shaping our journalism by becoming a member today.)
Published: undefined