Google closes net on hate speech publishers

Google made some subtle, but important changes to their policy on hate speech on April 26^th. The changes come after a challenging few weeks where Google faced criticism and advertiser walk-outs over the content on some parts of the Google Publisher Network (See Is Google really funding extremists?). The change shouldn’t come as a surprise, as Google SVP & Chief Business Officer Philip Schindler had already promised a tougher stance on “content that is attacking or harassing people based on their race, religion, gender or similar categories.”.

The changes come to Google’s policy on “Content that advocates against an individual, group, or organization“, more frequently referred to as the hate speech policy. The changes serve to clarify expectations, but also the widen the net on what would fall into that category. Before looking at the specifics let’s compare the old and new versions:

What the hate speech policy did say

Google believes strongly in the freedom of expression, but also recognizes the need to protect the quality of the AdSense network for users, advertisers and publishers.
Google ads may not be placed on pages that contain harassing or bullying content, or on content that incites hatred or promotes violence against individuals or groups based on race or ethnic origin, religion, disability, gender, age, veteran status or sexual orientation/gender identity. Additionally, Google ads may not appear on content that incited or advocated for harm against an individual or group.

What it says now

We believe strongly in freedom of expression, but we don’t permit monetization of dangerous or derogatory content. For this reason, Google ads may not be placed on pages containing content that:
• Threatens or advocates for harm on oneself or others;
• Harasses, intimidates or bullies an individual or group of individuals;
• Incites hatred against, promotes discrimination of, or disparages an individual group on the basis race or ethnic origin, religion, disability, age, nationality, veteran status, sexual orientate, gender, gender identity or other characteristic that is associated with systemic discrimination or marginalization.

What seems to have changed

Monetizing hate content now explicitly forbidden

The “need to protect the quality of the AdSense network” has now made way for the far clearer “we don’t permit monetization of dangerous or derogatory content”. That wording suggests strong enforcement to follow and the remaining changes make that enforcement far clearer and easier.

It’s page level

Google have been working more more granular page level enforcement for a while and publishers are seeing more pages blocked from serving ads over domain level notifications. We’ve talked about this change on the blog before, (see AdSense policy gets more granular). With regards to hate speech this particularly allows issues aroung hate-filled user-generated content to be effectively tackled.

Self-harm made the list

One particular criticism in the recent media storm was that Google was making money from ads shown on content that promotes self-harm. That hold in policy has been closed by banning use of Google ads on content that “Threatens or advocates for harm on oneself or others”.

Broadens negative behaviour caught by policy

The phrase “promotes violence against” is replaced by “Threatens or advocates for harm” and “incites hatred” broadened to be  “Incites hatred against, promotes discrimination of, or disparages”, presumably to tackle that narrow group of publishers who have become very adept at staying just the right side of the line without dialling down the hate.

nationality. other characteristic that is associated with systemic discrimination or marginalization.

Nationality gets protected

Nationality gets added to the list of characteristics that is protected and a new catch all “other characteristic that is associated with systemic discrimination or marginalization.” has been added to give some flexibility once people find a new group to hate on.

Summing up

The policy change itself seems reasonably straight forward, but like the recent changes on fake news, what really counts is how these policy changes are enforced.