Cognate's Trademark Blog

Copyright Bot Loses The Plot

Great report from Law.com about the Topple Track Bot although we have to say  we’re pretty pleased Ed Sheeran was a target !

 

Law.com reports

Topple Track Bot Goes Berserk with Bogus DMCA Notices

Source: https://www.law.com/2018/08/15/whats-next-law-bot-goes-berserk-plus-googles-watchful-eye/

Automation may be touted as a saving grace for the legal industry’s more mind-numbing tasks, but even robots make mistakes. Case in point: self-described “content protection service” Topple Track was called out for some big blunders last week over what appeared to be bogus Digital Millennium Copyright Act takedown notices. The bot targets included a New Yorker article, musician Ed Sheeran’s website, and the Electronic Frontier Foundation — which judging by a piece by EFF attorney Daniel Nazer, isn’t too happy with “copyright robots run amok.”

Essentially an automated way for musicians to send takedown notices, Topple Track was a member of Google’s Copyright Program, up until recently being pulled down. (The company said on its website that it’s “tightening some nuts and bolts”.) Quoting a page taken down by Topple Track’s parent company, Nazer writes that in practice, this means sending DMCA takedown requests to Google so URLs are deindexed from search and/or removed by website hosts.

In Nazer’s telling, the company is “a poster child for the failure of automated takedown processes,” particularly given the “scope of expression it has sought to delist” (EFF, for example, was targeted for a song called “My New Boy,” which might not even exist). And it isn’t just a send-and-forget obligation under the law; DMCA requires that senders affirm they have “a good faith belief” that a site is unlawfully using copyrighted material. In analyzing Topple Track’s bunk notifications, Nazer says EFF could find neither link nor “plausible claim of infringement” between the accused and the takedown notices.

Now, DMCA notifications are no laughing matter, and minor mishaps can have significant consequences. Nazer writes that many website owners remain oblivious to their URLs being targeted, or could be subject to a “copyright strike” – basically where a hosting site, say YouTube, takes down your content in response to a notification.

And while a considerable amount of questionable takedown requests can be chalked up to bot botches or “plain incompetence,” TorrentFreak reports that there’s been “an influx of more coordinated DMCA abuse.” The motive? Downranking competitors in Google search results, with malicious users posing as major entities under misleading names, such as ‘Walt Disney LTD.’ Whether sketchy or legit, those targeted can find recourse in issuing a counter-notice. However, as Stanford’s Daphne Keller and Annemarie Bridy write in a blog, “The number of potentially mistaken or malicious notices still vastly exceeds the number of counter-notices.”

This isn’t the only time the idea of scrubbing the web with bots has led to controversy. In September of last year, the European Commission issued a communication not only encouraging web platforms to employ measures for pulling illegal content, but “step up cooperation and investment in” automatic detection technology. In a fun twist of coincidence, TorrentFreak reports the trigger-happy Topple Track also targeted an article written by an EU Parliament Member,discussing web “censorship machines” and copyright, which was later nixed from Google’s search index. You can search for other bot victims here.

Thinking Ahead: Platforms are under greater pressure to keep watch over user content, but automated filtering isn’t proving to be such a hot option. Topple Track is one example — but seems unlikely to be the last.

Share This Blog Post:

Leave a Comment