Facebook's Content Moderation System Flagged and Removed Emergency Updates as Spam During Wildfires

During the 2024 wildfire season, Facebook's automated content moderation system began systematically flagging and removing emergency safety posts as spam, affecting at least 40 posts across 20 fires since June 2024. The posts contained critical information including evacuation orders, fire locations, and safety resources, often linking to official sources like Cal Fire, U.S. Forest Service, and sheriff departments. The removals affected volunteer fire safety coordinators, official fire departments, news stations, and disaster nonprofits across California and other states. Users received notifications stating their posts violated 'Community Standards on Spam' for trying to get engagement in a 'misleading way,' but were not notified when posts were actually removed. The issue particularly impacted rural communities like Lake County, California, where Facebook groups serve as primary information sources during emergencies due to limited official communication channels and spotty cell service. Some users had accounts suspended after multiple violations, and the Forest Service had an entire emergency information page suspended during the Leadville, Colorado fire. Facebook spokesperson Erin McPike stated the company was investigating the issue and working to address it, noting they were unaware of the problem until contacted by The Washington Post.