About 1.6 million views on YouTube out of every billion are from a video that violates its content policies, even a year ago, Alphabet Inc.’s Google-owned streaming service said in a new disclosure Tuesday.
The “viewing violation rate” (VVR) has dropped more than 70 percent since it was first tracked in the fourth quarter of 2017, YouTube said, and demonstrates its progress in blocking hate speech and other videos it considers dangerous before they go viral.
Critics have said that improper surveillance by YouTube and other social media companies allows false and hateful rhetoric to spread, fueling deadly violence like the attack on the United States Capitol in January.
YouTube’s VVR was stable for the last six measured quarters, according to the new data, which runs through 2020.
Jennifer O’Connor, YouTube’s chief product officer, told reporters that she expected the release of the estimate each quarter “will continue to hold us accountable.”
He said the rate, like other compliance data posted by YouTube, could fluctuate as its technology, rules and users evolve. For example, YouTube removed nearly 171,000 hate speech channels in the fourth quarter, three times more than in the previous period. He attributed the jump to improved sensing technology.
The VVR covers all policy violations and is derived from a video sample. It does not include comments on videos.
Facebook Inc publishes a similar estimate, but excludes harassment, spam, and other violations. In addition, Facebook has said that at least 15 million views out of 1 billion in the fourth quarter were for content that violates its rules against nudity and adult sexual activity, violent or graphic material and hate speech.
Countering criticism about the “rating” itself, Facebook said last year that it would hire an external auditor to evaluate its disclosures.
YouTube’s O’Connor declined Monday to commit to an external audit, but said he would “not rule it out.”