Net anonymity

Whether it's about Tunisia, Egypt - or as is currently the case, Syria - the global public largely gets its information from online sources. Citizen journalism - for example, amateur videos recorded by anonymous participants and uploaded to platforms such as YouTube - has been playing an important role in getting information about police or military repression out into the world.

This documentation of people's political movements, important as it may be, also has a downside: those portrayed may become subject to repression by the state or other actors seeking out the leaders of the movement. It's to address this possibly deadly danger that Google has developed a new tool for its video platform YouTube.

Risk for activists

The downsides and upsides of citizen journalism have long been debated - what such coverage lacks in journalistic standards could be made up for by authenticity and immediacy.

But this very immediacy also poses a danger. The fact that videos are often uploaded directly after being recorded, usually with little or no editing, can have serious consequences for those portrayed.

'Among the first'

This is exactly the scenario YouTube had in mind when it developed its face-blurring function, the company wrote in its official blog.

"As citizens continue to play a critical role in supplying news and human rights footage from around the world, YouTube is committed to creating even better tools to help them," the blog post stated.

YouTube noted that it was "among the first" to offer such identity-protection measures.

The feature can be found by going into "Video Manager," clicking the "Edit" button to the right of the video, then selecting "Enhancements" from the top menu. "Blur all faces" is the only option under the "Additional features" button that shows up beneath the video to be edited.

After hitting "Apply," the edited video can be previewed alongside the original. The original can then be deleted - "very quickly," as YouTube's Jessica Mason was quoted as saying on the news site Ars Technica.

Imperfect results

But responsible activists who use the face-blurring function may end up wanting to immediately erase the resulting video - it's far from perfect, as even YouTube admits.

"This is emerging technology, which means it sometimes has difficulty detecting faces depending on the angle, lighting, obstructions and video quality," the blog post states.

In its own test of the feature, Ars Technica reported that with poor images, "the blur tool would blur out large swaths of the image, in a hilarious (or frustrating) case of over-correction."

But the site also reported that the company was working on improving the technology, which relies on a face-recognition algorithm.

Face-recognition continues in controversy

The technology brings dangers with it: such algorithms can not only establish where in an image a face is, but are very close to being able to identify individuals in photos - this feature has been "built into" Facebook, despite user resistance to social networks acting as Big Brother.

Google has also made progress in detecting the presence of individuals in videos; a patent for the technology became known at the beginning of July.

So along with praise for YouTube's new face-blurring software, privacy and human rights advocates continue to fear that the related face-recognition technology could be exploited by repressive regimes.

The issue recently garnered attention at a hearing in the United States Senate, where Senator Al Franken raised concerns over the software.

"I fear that without further protections, this technology could be used on unsuspecting civilians innocent of any crime," Franken, a Democrat, testified.

When Google's new facial recognition software comes out, it could end up saving intelligence agencies and state spies a lot of work.

Author: Michael Gessat / sad
Editor: Michael Lawton