Google says it took down 10m IP-infringing ads in 2017

Google says it removed more than 10m adverts suspected of infringing copyright last year as part of a crackdown on online piracy.

The company also says it has blocked 3bn URLs from its search engine for infringing copyright, since launching a submission tool that can be used copyright owners and their agents to flag up problem sites.

The claims are made in a new report that lays out Google’s approach to “ensuring creators and artists have a way to share and make money from their content – and preventing the flow of money to those who seek to pirate that content,” according to a blog post by Cedric Manara, Google’s head of copyright.

Google says the guiding principles in this area are to create better alternatives to piracy that encourage users to pay for content, pointing to its Google Play Music and YouTube platforms, and making it harder for infringers to make money by ejecting rogue sites from its advertising and payment services.

It’s also been working to make the removal process for copyright infringement quicker and more efficient, root out fabricated infringement allegations use for censorship or anti-competitive reasons, and making its processes more transparent.

“These efforts appear to be having an effect: around the world, online piracy has been decreasing, while spending on legitimate content is rising across content categories,” says Manara. “Today, our services are generating more revenue for creators and rights holders, connecting more people with the content they love, and doing more to fight back against online piracy than ever before.”

The report suggests that Google has generated $3bn in revenue for video rightsholders on YouTube, as well as paying another $6bn in ad revenue to the music industry. It’s also spent more than $100m on its recently implemented (and sometimes controversial) Content ID copyright management system, which automatically scans videos for audio and video content deemed infringing. 98 per cent of copyright claims on YouTube in 2017 were made through Content ID.

“Content ID can now catch efforts to evade detection like changing a video’s aspect ratio, flipping images horizontally, and speeding up or slowing down the audio,” says the report, which claims that fewer than 1 per cent of Content ID claims are disputed. The technology can also “detect copyrighted melodies, video, and audio, helping identify cover performances, remixes, or reuploads they may want to claim, track, or remove from YouTube.”

Turning to search, Google says that last year it received notifications from content owners of 882m URLs, a 9 per cent reduction on the prior year, with 95 per cent of them removed leaving around 54m rejected  on the grounds they were “incomplete, mistaken, or abusive.” The decline in notifications reverses a relentless upward trend in recent years, it says.

“We’re proud of the progress this report represents,” says Manara. “Through continued innovation and partnership, we’re committed to curtailing infringement by bad actors while empowering the creative communities who make many of the things we love about the internet today.”