What YouTube has said:
- It has removed a lot more content: YouTube claims that its June policy update against hate speech and supremacist content is working, it has removed 17,000 channels coming up to over 100,000 videos, both of which are 5x that of the last quarter. It has also removed over 500 million comments, which has doubled over the last quarter “in part due to a large increase in hate speech removals”.
- Automation is working: YouTube also says that automated flagging is working, and that 87% of the 9 million videos removed in Q2 were first flagged by its automated systems. 80% of these were removed before a single person saw them. An update to its spam detection system in Q2 of 2019 led to a 50% increase in the number of channels removed for violating spam policies.
- Violative videos are removed before people see them: It is getting better at removing videos before they get views. 30,000 videos removed for hate speech over the last month got 3% of the views that knitting videos did (in the same period).
- 10,000 people across Google working with detecting, reviewing, and removing content.
- Views on violative videos reduce by 80% before removal: YouTube says that over the last 18 months, changes it made to its content policies and removal practices have reduced views on videos that are later removed for violating its policies by 80%.
What was the June policy change? YouTube said it would be much more stern about hate speech and supremacist content, and escalated policy from having a “tough stance” towards such content to prohibiting it via removals. It also said spread of borderline content and harmful misinformation — such as videos which claims that the Earth if flat — will see lowered distribution. Go deeper here.
YouTube is also working on creator-on-creator harassment: YouTube’s policy team is currently working on updating its harassment policy, including and most notably to address creator-on-creator harassment (this was first announced in April 2019). YouTube received backlash earlier this year when it stated that Steven Crowder — a US conservative commentator with 4 million YouTube subscribers — had not violated YouTube violating for having used racial language and homophobic slurs to harass Carlos Maza, a Vox journalist, for nearly two years.
Strategic timing? Potentially, as YouTube made this public just days after it agreed to pay between $150 million to $200 million to the Federal Trades Commission (FTC) for violation of the Children’s Online Privacy Protection Act (COPPA) in the US.