Automated Moderation Is Here to Stay

According to the Electronic Frontier Foundation, automated content moderation has quietly graduated from emergency measure to permanent fixture of how online platforms police speech. In a new analysis, the digital rights group revisits a warning it issued six years ago, at the start of the pandemic, that stopgaps adopted in a crisis tend to outlast the crisis. That prediction, they argue, has held. The Foundation traces the arc from early spam filters and hash-matching tools to Facebook's 2017 turn toward AI for flagging extremist content, and notes that by 2018 the company claimed AI caught ninety-nine percent of ISIS and Al Qaeda material before any human saw it. The upside, EFF acknowledges, is sparing human moderators from grueling and traumatic work. But it points to well-documented costs: over-removal, bias baked into training data, and the erasure of minority languages and marginalized voices, citing a 2025 joint declaration from United Nations and other human rights bodies. The real question now, the group says, is not whether platforms will automate moderation, but under what conditions, transparency, and accountability they should be allowed to.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton