Breaking Bug on X Causes Posts to be Labeled as “Sensitive Media”
A recently discovered bug on X, formerly known as Twitter, caused chaos over the weekend. Many posts were incorrectly flagged as “Sensitive Media,” despite the company’s efforts to make their platform more appealing to advertisers. However, the X Safety account has assured users that the glitch has been fixed and they are working to remove the labels from affected posts.
X uses the label “Sensitive Media” for content that may be deemed inappropriate, such as violence or nudity. To prevent these types of media from appearing on their platform, X asks users to adjust their settings accordingly. Additionally, there is an option to add a warning for sensitive content on photos and videos across X on iOS, Android, and the web. This can include categories such as nudity, violence, or just “sensitive” in general, all of which require an extra click or tap to view.
While X typically relies on automation and human review from their trust and safety team to address reports of sensitive content, this time it seems the issue may have been caused by a spam bot. In a post by owner Elon Musk, he mentions that a “combination of automation and human review” is used for these cases. However, in the same post, he also suggests that a spam bot may have been responsible for the glitch, which contradicts the official statement from the X Safety team.
Musk tweeted on Sunday, “An X spam/scam bot accidentally flagged many legitimate accounts today. This is being fixed.” He later reposted the official statement from the X safety team, referring to the issue as a bug. The team later confirmed that all affected posts have been corrected and the erroneous labels have been removed.
This latest hiccup is just one in a series of missteps X has faced since Musk’s takeover. After facing a mass exodus of advertisers due to concerns over anti-semitic content, Musk notoriously told them to “go f*ck yourself.” As a result, X has been seeking out smaller advertisers while they work on implementing AI and peer-to-peer payments in 2024.
Unfortunately, this bug may have been intensified by X‘s recent downsizing, particularly in their trust and safety team – the group typically responsible for reviewing accounts for spam and sensitive content.
Bots flagging accounts is not the only issue X has dealt with in recent months. A search for the phrase “I’m sorry, I cannot provide a response as it goes against OpenAI’s use case policy” revealed numerous automated accounts posing as genuine users, with many even paying for X Premium. Musk had previously thought that introducing a small fee would help weed out spam, but this demonstrates that some bots are willing to pay to appear human. Last summer, the company also admitted to a Verified spammer problem and implemented new DM settings to move messages from these accounts out of users’ inboxes.
Since Musk’s takeover, X has experienced other major bugs, including one in August 2023 that caused issues with native images and links, and a global outage just last month.
