NASA’s Mars Rover Crashed Into a DMCA Takedown
NASA’s livestream coverage of the Curiosity rover’s landing on Mars was was practically as flawless as the landing itself, a refreshing alternative to all that troubled Olympics coverage. The broadcast – full of suspense, lucky peanut-eating, and ecstatic congratulations – was slow and hard to reach at times, but NASA servers never failed. Along with burnishing its online publicity credentials, NASA had prepared for a global audience of millions
But NASA couldn’t prepare for everything. An hour or so after Curiosity’s 1.31 a.m. EST landing in Gale Crater, I noticed that the space agency’s main YouTube channel had posted a 13-minute excerpt of the stream. Its title was in an uncharacteristic but completely justified all caps: “NASA LANDS CAR-SIZE ROVER BESIDE MARTIAN MOUNTAIN.”
When I returned to the page ten minutes later, I saw this:
Stop the band. The video was gone, replaced with an alien message: “This video contains content from Scripps Local News, who has blocked it on copyright grounds. Sorry about that.” That is to say, a NASA-made public domain video posted on NASA’s official YouTube channel, documenting the landing of a $2.5 billion Mars rover mission paid for with public taxpayer money, was blocked by YouTube because of a copyright claim by a private news service.
-While it’s not exactly clear what happened here, it’s possible that an affiliate of the news service had uploaded a video that contained material from NASA’s own stream before NASA itself could upload it. YouTube’s DMCA bots wrongly assumed the video belonged to Scripps. That, or a human at Scripps filed a copyright claim.- NASA, which has a powerful presence on YouTube, likely contacted Google directly for help getting the video reinstated.
Within hours, the problem was fixed (and the title switched back to a calmer regular title case — see the video below). But it was still a disappointing blip in an otherwise exceptional moment for humanity, what President Obama called “an unprecedented feat of technology.” It was also an interesting little object lesson in what’s still wrong with online copyright enforcement right now.
On Monday afternoon, a spokesperson for E.W. Scripps Company, owner of the news service, emailed Motherboard a statement apologizing for the accidental takedown. “We apologize for the temporary inconvenience experienced when trying to upload and view a NASA clip early Monday morning,” wrote Michele Roberts. “We made a mistake. We reacted as quickly as possible to make the video viewable again, and we’ve adjusted our workflow processes to remedy the situation in future.”
This isn’t the first time that a claim by the company, Scripps News Service, has grounded a NASA YouTube video: it happened in April, with a video of one of NASA’s Space Shuttles being flown atop a 747. According to Bob Jacobs, NASA’s Deputy Associate Administrator for Communications, such blocks happen once a month, and tend to be more common with popular videos.
“Everything from imagery to music gets flagged,” Jacobs told me this morning. “We’ve been working with You Tube in an effort to stop the automatic disabling of videos. So far, it hasn’t helped much.
“The good thing about automation is that you don’t have to involve real people to make decisions. The bad thing about automation is that you don’t have to involve real people to make decisions.”
This isn’t a phenomenon limited to NASA’s videos: YouTube’s sophisticated copyright policing system is riddled with problems that make it prone to abuse and accidents, and that often make it hard to unblock videos that have wrongly been removed.
How YouTube’s Copyright Robots Work (or Don’t)
YouTube will block or censor content for one of three reasons: if a video violates the site’s terms of service, if its content is automatically found to match copyrighted content, or if it receives a request from a copyright owner to remove a pirated video. The blockage of pirated content is is governed by the Digital Millenium Copyright Act (DMCA) of 1998, the Congressional law that governs how websites deal with intellectual property violations and that stood to be replaced in part by this year’s Stop Online Piracy Act.
Content ID, YouTube’s automated copyright monitor, was meant to be the site’s secret weapon in its fight to stay legal, and make some sense – or cents – out of the video chaos: by algorithmically matching content, robots can, ideally, keep track of which videos contain copyrighted material. If owners (like record companies) want to make ad revenue from creative reworkings – or wholesale copies – of their content, they can do so while allowing uploaders to keep their copyrighted videos online. And the system also allows copyright owners to automatically block videos that appear to infringe their copyright.
In its early days, YouTube made much of its revenue from illegal uploads; now content companies are taking Google to court to ensure the system is as strict as it can be.
But three years after Content ID launched, YouTube’s copyright policing regime, like the general copyright system it tries to enforce, is still flawed. It mistakenly blocks videos that shouldn’t be blocked, and encourages excessive claims. A few months ago, one user’s video of foraging for salad in a field was taken down because the media company Rumblefish claimed to own the soundtrack of singing birds.
YouTube’s system is also heavily biased in favor of claimants, and a system that is increasingly controlling of content that has serious educational or scientific value, or arguably falls under “fair use” provisions. Claims of fair use of video content are immaterial to the Content ID or DMCA takedown system. Creative remixes are easy targets, as are videos of teenagers singing Christmas songs. As I discovered last year, many of Martin Luther King, Jr.‘s speeches are no longer available on YouTube thanks to automatic and manual copyright claims by the owner of King’s speeches, the British music giant EMI Publishing. This despite the fact that YouTube is still a haven for illegal and uncontested uploads of millions of hours of Hollywood and music material.
YouTube’s first line of defense against bogus claims? A warning that says, “Don’t make false claims!”
And because anyone can claim a YouTube video belongs to them, YouTube’s system allows cheaters to direct traffic (and ad revenue) to their uploads. (Some preferred companies, like Universal Music Group, can even block videos immediately, without filing a claim.) Of course, hiring enough staff to pre-screen the 72 hours of video uploaded to the site every minute for possible copyright infringement would cost Google something close to its ad revenue for 2011, about $37 billion, one site estimated. But YouTube’s first line of defense against bogus claims is little more than a boldface notice on its copyright claims page that reads, simply, “Don’t make false claims!”
If the system made it easy for users who were falsely accused of copyright violations to reinstate their videos, abuses and failures of the system might be easier to prevent. But the process for reinstating a video is byzantine and difficult, and puts the burden of proof on the accused, not the accuser.
“We spend too much time going through the administrative process to clear videos slapped with needless copyright claims,” says NASA’s Bob Jacobs. “YouTube seems to be missing a ‘common sense’ button to its processes, especially when it involves public domain material paid for by the American taxpayer.”
According to the DMCA, if a user disputes a claim from a copyright holder, YouTube should make a video available again, at least until the copyright holder files a second claim to take the disputed video offline again. Instead, YouTube requires the alleged violator to submit a signed counter-claim, under penalty of perjury. (There’s no such penalty for those claiming violations to begin with.) YouTube forwards the claim to the supposed copyright owner and waits ten days for a response. “If we do not receive such notification, we may reinstate the material,” says YouTube, emphasis mine.
“Guys, our video is back online!”
NASA fans were incensed at the brief takedown. “If this isn’t a legitimate copyright claim, then they should be prosecuted to the full extent of the law,” wrote user eluhn on a YouTube forum. “Takedown requests should have to be submitted under the penalty of perjury, at the very least,” wrote a British commenter on a different forum. Another quipped, “I doubt any local news station got Cameras on Mars before NASA to watch the landing.”
Jacobs, who manages the agency’s main YouTube channel, isn’t just concerned about accidents that result from automated systems like Content ID, but about the rise of fraudulent blocks by uploaders wanting to claim NASA content as their own. “There seems to be few consequences for companies that engage in such activities, which often include legitimate news organizations. We do agree that people who make false copyright claims against our material should be held accountable, regardless of their automated systems.”
Robots, which are still built by humans, fail humans all the time: remember how Frankenstein algorithms burned through $10 million a minute on Wall Street last week. Good thing then that we have the video in question to remind us how exciting it can be when our robots don’t fail.