Facebook is rolling out a broad set of changes aimed at misinformation, harmful communities, impersonation and low-quality material across Facebook, Instagram and Messenger. The company presented the work as an expansion of an integrity strategy it has used since 2016: remove content that violates the rules, reduce the distribution of problematic material that remains allowed, and inform people before they decide what to trust or share.
The three-part framework recognizes that not every difficult post belongs in the same category. Hate speech, terrorist material and other violations of Facebook’s Community Standards can be removed. Clickbait or a sensational claim may not cross that line, but Facebook can rank it lower. Additional labels and publisher details are intended to help users judge material that remains visible.
Remove: clearer rules and closer attention to groups
Facebook is adding a section to its Community Standards site where people can follow policy changes month by month. The company says it revises rules in response to new behavior, feedback from reviewers and outside experts, and it wants to make substantive changes easier to track.
Groups will receive more scrutiny as well. When Facebook considers whether a group violates its standards, violations by administrators and moderators — including member posts they approve — will become a stronger signal. A new Group Quality feature will show administrators content removed or flagged for common violations and identify false news found in the group.
Reduce: fact-checking and a new ranking signal
Facebook is expanding the Associated Press role in its third-party fact-checking program. AP will review misleading videos and Spanish-language content in the United States in addition to other claims. Groups whose members repeatedly share stories rated false by independent fact-checkers can have their overall News Feed distribution reduced.
A new signal called Click-Gap compares a website’s popularity on Facebook with its authority across the wider web. A domain that receives an unusually large number of Facebook clicks but few links elsewhere may be succeeding in News Feed without having earned comparable recognition outside the platform. Facebook says that mismatch can indicate low-quality content and will become one input in ranking, not an automatic verdict by itself.
Instagram is applying a similar reduction principle to recommendations. Posts that are sexually suggestive or otherwise inappropriate may remain available to followers when they do not violate Community Guidelines, but they can be excluded from Explore and hashtag pages. The distinction lets Instagram limit amplification without treating every borderline post as prohibited.
Inform: more context on posts, pages and messages
Facebook is testing its News Feed Context Button on images reviewed by fact-checkers and adding Trust Indicators for English- and Spanish-language publishers. Those indicators can describe a newsroom’s ownership, funding, corrections policy, ethics and editorial staff. The Page Quality tab will also gain more information, including a page’s status in relation to clickbait.
Messenger is receiving tools intended to reduce deception and unwanted contact. Facebook’s Verified Badge is coming to Messenger to make impersonation harder, while updated messaging settings and blocking controls offer users more choice over who can reach them. A Forward Indicator and Context Button can show that a message was passed along and provide background about a linked article.
No ranking signal or fact-checking partnership can settle every dispute over truth, and Facebook says it will consult academics, journalists and civil-society groups as the systems develop. The April changes nevertheless show a shift from reacting only after a violation is reported toward controlling how questionable material gains reach throughout the company’s apps.
Recent Comments