Skip to main content
  1. Home
  2. Social Media
  3. News

Twitter is tweaking its algorithms after a bug hid results for LGBT keywords

Add as a preferred source on Google

After users pointed out searches for LGBT terms like #bisexual and #gay failed to turn up any search results last week, Twitter is updating an outdated search algorithm designed to hide sensitive topics. On Monday, November 6, Twitter apologized for the error and said that the missing search results were part of a technical bug.

Twitter says that sensitive media is collapsed in the search results and when all the results are collapsed, the search will turn up a “no results” message. The search terms themselves were not meant to categorize a tweet as sensitive material, Twitter says, but only to alert the program to look for explicit content. A list of sensitive words, such as the LGBT terms that weren’t showing up in the search results, triggers a look into the tweet containing one of those words for explicit content, but the terms themselves do not automatically mark tweets as explicit, Twitter said.

Recommended Videos

“As outlined in our media policy,” the company said via a tweet series, “media that may be considered sensitive is collapsed in places such as search results, meaning that images and videos would be presented as a link, not automatically updated. One of the signals we use to identify sensitive media is a list of terms that frequently appear alongside adult content. Many of these words on the list are not inherently explicit, which is why they must be used alongside other signals to determine if content is sensitive.”

While the lists were not supposed to mark content as explicit based on those terms alone, Twitter says an error allowed the tweets to be categorized solely on the text on not the context of those tweets. One of the lists, Twitter says, was also outdated and “included terms that are primarily used in non-sensitive contexts.”

The social media platform says that it has removed the terms that should not have been included on that list and is continuing to work on the issue to fully correct the bug.

Twitter is already revising its community guidelines with stricter policies on hate speech and harassment, changes sparked by the #WomenBoycotTwitter protest that will roll out over the next few months.

Hillary K. Grigonis
Hillary never planned on becoming a photographer—and then she was handed a camera at her first writing job and she's been…
X’s creator payouts are changing, and original content is the new currency
The platform is replacing Revenue Sharing with a program designed to reward creators who actually make something.
X Twitter on iPhone

X is changing the rules for creators yet again. This time, the platform wants to put less emphasis on simply generating engagement and more on actually making something original. X is replacing its controversial Creator Revenue Sharing program with a new initiative called Original Content Rewards, which will officially launch on September 8, 2026. The existing Revenue Sharing program will continue paying participating creators through September 7, after which the new system takes over.

X wants to reward creators, not engagement farmers

Read more
Snapchat’s Spotlight algorithm now favors human-made videos over AI-generated ones
Snapchat is taking a firm stance against AI slop.
Snapchat-App-Store-open-on-iPhone

If you've been uploading fully AI-generated videos on Snapchat in hopes of earning rewards, it might be time to rethink your strategy. Snapchat has announced changes to Spotlight that prioritize authentic, original creativity over content created entirely by artificial intelligence. The company says its recommendation system will now favor videos made by real creators, while fully AI-generated submissions will no longer qualify for monetization.

So what's changing on Spotlight?

Read more
LinkedIn is crowdsourcing its fight against AI slop
The company is rolling out a new button that lets you flag posts that seem AI generated.
LinkedIn seems like AI slop button

LinkedIn has a serious AI slop problem, and it's now turning to its own users to help fix it. The platform is rolling out a new button that lets users flag posts they suspect were generated by AI. It plans to use that feedback to fine-tune the system that decides how much reach a post gets outside a user's own network, based on how AI-generated it appears.

The announcement comes shortly after an analysis by Pangram found that more than forty percent of long-form LinkedIn posts are fully AI-generated, and it suggests that the reach-trimming measures the company rolled out earlier this year haven't been enough on their own.

Read more