Bluesky's 2024 moderation report shows how quickly harmful content grew as new users flocked in

Bluesky experienced explosive growth last year, particularly toward the end, necessitating that the platform ramp up its moderation efforts. In its recently released moderation report for 2024, Bluesky said it grew by about 23 million users, jumping from 2.9 million users to nearly 26 million. And, its moderators received 17 times the number of user reports they got in 2023 — 6.48 million in 2024 compared to 358,000 the previous year. 

The bulk of these reports were related to “harassment, trolling or intolerance,” spam and misleading content (including impersonation and misinformation). The presence of accounts posing as other people has been a known issue in the wake of Bluesky’s popularity spike, and the platform updated its impersonation policy in November with a “more aggressive” approach in an attempt to crack down on it. At the time, it said it had quadrupled its moderation team. The new report says Bluesky’s moderation team has grown to about 100, and hiring is ongoing. “Some moderators specialize in particular policy areas, such as dedicated agents for child safety,” it notes.

Other categories Bluesky says it received a lot of reports about include “illegal and urgent issues” and unwanted sexual content. There were also 726,000 reports marked as “other.” Bluesky says it complied with 146 requests from “law enforcement, governments, legal firms” out of a total of 238 last year.

The platform plans on making some changes to the way reports and appeals are handled this year that it says will “streamline user communication,” like providing users with updates about actions it has taken on content they’ve reported and, further down the line, letting users appeal takedown decisions directly in the app. Moderators took down 66,308 accounts in 2024, while its automated systems took down 35,842 spam and bot profiles. “Looking ahead to 2025, we're investing in stronger proactive detection systems to complement user reporting, as a growing network needs multiple detection methods to rapidly identify and address harmful content,” Bluesky says. 

This article originally appeared on Engadget at https://www.engadget.com/social-media/blueskys-2024-moderation-report-shows-how-quickly-harmful-content-grew-as-new-users-flocked-in-000149354.html?src=rss https://www.engadget.com/social-media/blueskys-2024-moderation-report-shows-how-quickly-harmful-content-grew-as-new-users-flocked-in-000149354.html?src=rss
Établi 16d | 20 janv. 2025 à 00:40:15


Connectez-vous pour ajouter un commentaire

Autres messages de ce groupe

Netflix scuttles plans to add six previously announced games to its service

Netflix has been revamping its games division in recent months, including making

4 févr. 2025 à 20:50:05 | Engadget
Cruise lays off half its staff after GM sunsets robotaxi program

Autonomous vehicle company Cruise is laying off around half of its workforce,

4 févr. 2025 à 20:50:04 | Engadget
Government workers sue over potentially illegal DOGE server

Federal employees are suing to disconnect a server, reportedly operated by associates of Elon Musk, from the US Office of Personnel Management. A

4 févr. 2025 à 20:50:02 | Engadget
Adobe's Acrobat AI Assistant can now assess contracts for you

Adobe has updated the Acrobat AI Assistant, giving it the ability to un

4 févr. 2025 à 18:30:40 | Engadget
WhatsApp brings image and voice inputs to its ChatGPT integration

The tech sector’s ongoing effort to force-feed generative AI features into widely used services continues with updates to

4 févr. 2025 à 18:30:39 | Engadget
Our favorite budget Roomba is half off right now

I really hate vacuuming and put it off as long as possible. But, my budget isn't very robot vacuum friendly, so, alas I have no choice. However, that might change, as our

4 févr. 2025 à 18:30:38 | Engadget
Verizon is adding Google One AI Premium as a perk for some customers

Verizon’s myPlan wireless and myHome internet subscriptions allow users to choose from a selection of

4 févr. 2025 à 18:30:37 | Engadget