The Tech Transparency Project first noticed the bug and believes that it has been in effect since at least June. “What we’ve seen with the related hashtags with Biden is that there is so much vitriol, that it appears to be benefitting President Trump,” director of the Tech Transparency Project, Katie Paul, told Buzzfeed on Wednesday.
An Instagram spokesperson characterized the issue as a ‘bug’ and pointed out that other non-political hashtags such as #menshair and #artofdrinks were similarly impacted. “A technical error caused a number of hashtags to not show related hashtags,” the spokesperson said in a Tuesday statement. “We've disabled this feature while we investigate.”
The Biden campaign declined to comment however a spokesperson for the Trump campaign “lambasted the bug as an example of alleged bias against conservatives on social media platforms,” according to Buzzfeed.
This is far from the first time that conservatives have railed against what they perceive as bias, even when actively benefitting from it. Just last August Facebook commissioned a third-party audit of its platform in the wake of three years of congressional Republican baseless wingeing about how the social media company inherently promoted liberal content. This, of course despite the fact that the site’s Top 10 daily posts, as measured by Facebook itself, are overwhelmingly produced by fringe right-wing outlets like Ben Shapiro and Sean Hannity.
If you want to see how toxic the related hashtags feature can be, look how it spread conspiracy theory hashtags when associated with #ChrissyTeigen. This is what Instagram’s algorithm thinks people want to see. pic.twitter.com/CnoCA9CKKz — Ryan Mac 🙃 (@RMac18) August 5, 2020
To be fair, however, Instagram’s recommendation algorithm is a bit garbage to start with. As Buzzfeed reporter Ryan Mac points out above, the platform routinely pops worrisome, conspiracy theory-related hashtags for searches of famous figures such as Chrissy Teigen.
petripeeduhpedro on August 6th, 2020 at 00:59 UTC »
Here's what engadget's article links to as their source.
Most of what the disparity (and there's definitely a difference shown in the article) is linked to is the appearance of related hashtags. Normally, IG shows related hashtags when you're looking at a hashtag. They were doing this for Biden but not for Trump hashtags. So things like #sneakyjoebiden could slip in as related to #Biden, but nothing like this was showing up for Trump.
Personally, I think there could be malicious intent behind it, but I also think that due to online propaganda communities historically popping up for Trump, IG may have disabled related hashtags as a brute method to avoid spreading misinformation or hate speech. And possibly they didn't think to do that for Biden because of either a lack of foresight for the cutthroat nature of politics or due to online platforms' tendency to react only after something bad has happened rather than before (as in rollercoasters are fixed only after someone gets hurt).
From reading into it a bit, I don't think we know enough to make a claim either way. I do think that regardless, this story highlights how damaging the lack of regulation for these platforms is. These platforms (twitter, facebook, instagram, youtube, etc.) have a history of questionable "bugs" and even censorship - we need to be able to hold them accountable or give them legal expectations to meet.
Conquestadore on August 5th, 2020 at 20:41 UTC »
Reminds me of the sub-plot in house of cards where one presidential candidate was manipulating search results in the equivalent of Google to try to come out on top.
The show did get some things right.
dickmove2020 on August 5th, 2020 at 18:48 UTC »
Zuckerberg working to get Trump elected....again.