Genius Accuses Google of Copying Its Lyrics Data
Jon Fingas (via Nick Heer):
Have you found yourself using Google’s lyrics results more than visiting individual lyrics sites? You’re not alone -- and Genius thinks underhanded tactics are involved. The company has accused Google of not only hurting its traffic with its lyrics cards, but of sometimes grabbing Genius’ lyrics verbatim. The evidence is in the apostrophes, Genius said. It purposefully alternated between straight and curved apostrophes as a form of watermarking (they typically spell out “red handed” in Morse code), and there were reportedly over 100 instances where Google’s lyrics included those exact apostrophes.
The lyrics displayed in the information boxes and in Knowledge Panels on Google Search are licensed from a variety of sources and are not scraped from sites on the web. We take data quality and creator rights very seriously, and hold our licensing partners accountable to the terms of our agreement.
See also: Jack Nicas (via Hacker News):
Online-reviews firm Yelp Inc. alleged that Google is breaking a promise it made as part of a 2012 regulatory settlement to not scrape content from certain third-party sites including Yelp, escalating its yearslong battle against the search giant.
Previously: Xerox Scanners and Photocopiers Randomly Alter Numbers.
Update (2019-06-18): Janko Roettgers (via Joe Rosensteel):
LyricFind said that it had been contacted by Genius in the past, and offered to remove any content that the company took issue with — but that Genius didn’t respond to this offer. The company also aimed to put Genius’ claims in perspective:
“Genius claims, and the WSJ repeated, that there are 100 lyrics from Genius in our database. To put this into perspective, our database currently contains nearly 1.5 million lyrics,” it wrote. “The scale of Genius’ claims is minuscule and clearly not systemic.”
Here’s my uninformed guess on what’s going on with Genius/Google: publishers may be scraping Genius themselves (since they often don’t get lyrics directly from writers) and delivering that data to the lyric licensors themselves.
To help make it clearer where the lyrics come from, we’ll soon include attribution to the third party providing the digital lyrics text. We will continue to take an approach that respects and compensates rights-holders, and ensures that music publishers and songwriters are paid for their work.
Update (2020-01-02): Aaron Sankin:
After Genius’s REDHANDED plan was made public, Google said it fixed the issue. Genius secretly snuck another “watermark” into its lyrics and found that Google was still stealing it content.
Update (2020-08-17): Devin Coldewey (via Hacker News):
A state court has dismissed a high-profile case showing unsportsmanlike conduct by Google, which was caught red-handed using lyrics obviously scraped from Genius.
[…]
The problem is this: Genius isn’t the copyright holder for these lyrics, it just licenses them itself. Its allegations against Google, Judge Margo Brodie of the Eastern District of New York determined, amount to copyright violations, in nature if not in name, and copyright is outside Brodie’s jurisdiction.
