Spotify has bought a company it says will help it detect and address harmful content on the platform. Kinzen uses machine learning and human expertise to analyze possibly harmful content and hate speech across multiple languages, Spotify said in a statement. It added that Kinzen will “help us more effectively deliver a safe, enjoyable experience on our platform around the world” and that the company’s tech is especially suited to podcasts and other audio formats.
Looking for signs of harmful audio content across hundreds of languages is a difficult task, particularly since there may be nuances that certain systems and non-native speakers may not pick up on. As such, Spotify says Kinzen will help it to “better understand the abuse landscape and identify emerging threats on the platform.”
The two companies have actually been working together since 2020, with the aim of preventing misinformation in election-related content. They forged their partnership before Joe Rogan started spreading COVID-19 vaccine misinformation on his Spotify-exclusive show, which is said to be the most-listened-to podcast on the planet.
There was a significant backlash against Rogan and Spotify earlier this year. Hundreds of doctors and scientists urged Spotify to enact a policy to stem the spread of misinformation. Weeks later, Spotify started adding a content advisory to podcast episodes that included discussions about COVID-19 and made its content guidelines public.
Several prominent artists pulled their music from Spotify in protest against Rogan, including Neil Young and Joni Mitchell. The host and editor of Spotify-owned podcast Science VS also pledged to stop making episodes (except ones in which it fact-checked Rogan and countered misinformation on the platform) unless the company did more to stem the flow of harmful falsehoods.
It may be the case that Spotify sees employing Kinzen’s tech as a means to help it avoid similar PR catastrophes in the future. However, it’s unclear how effective it will be in preventing the spread of misinformation. Were harmful content to be featured on a popular podcast, it may gain traction on social media and in press coverage before Spotify can act. Still, at least it’s trying to do more to stamp out misinformation.