The contentious debate over facial recognition software – and now mask recognition software – rages on as developers race to pioneer the best algorithms, despite substantial criticism from privacy advocates.
With 34 states in the U.S., plus the District of Columbia, enforcing mask mandates, recognition software has begun populating the markets. In a National Geographic article last week, Akash Takyar of LeewayHertz in San Francisco said that his company’s recognition software has been deployed covertly in hotels and restaurants across the U.S. and Europe as a tool employers use to make sure their staff remain mask compliant. An airport in the U.S. is reportedly testing the technology on site as well. The United States does not have a federal law governing data privacy, so the question as to whether recognition software crosses the line into privacy infringement or not remains undecided.
Proponents of mask recognition software claim it is not anti-privacy because it does not identify specific people. Facial recognition software, like the kind that allows iPhones to identify their owners’ faces, relies on a “training set,” an initial face it’s trained to identify. Mask recognition software is intended only to identify if someone is wearing a mask or not. Avast Security Evangelist Luis Corrons thinks the benefits outweigh any potential problems. “When we are in the middle of a pandemic bigger than any of us have ever known,” he said, “using technology to avoid dangerous behaviors like not wearing a mask in closed places seems reasonable. We are not talking about invading people’s privacy, this is about making sure that anyone potentially risking the health of others can be noticed so action can be taken.”
Opponents cite the software as imperfect in that it would allegedly capture much more data than simply whether or not someone was wearing a mask, and multiple cases of erroneous facial recognition attempts have proven that certain skin tones can confuse the algorithms. Moreover, some opponents worry that it would mark the beginning of more widespread surveillance which, left unchecked, would become normalized when the pandemic is over.
YouTube hit with £2.5B lawsuit over children’s data
London-based nonprofit Foxglove has taken representative action against YouTube, seeking £2.5 billion in damages for allegedly breaking U.K. and European data protection laws by targeting up to five million children under the age of 13 with addictive programming and data harvesting. In its press release, Foxglove maintained that Google-owned YouTube pitched itself to toymakers Hasbro and Mattel as “the favorite website of kids 2-12.” Foxglove stated that if the case succeeds, “millions of British households whose kids watch YouTube may be owed hundreds of pounds.” In a comment to TechCrunch, a spokesperson for YouTube claimed the website was not intended for children under the age of 13, referring them instead to the YouTube Kids app, developed in 2015.
TikTok pushes to settle child privacy lawsuit ahead of sale
As the Oracle acquisition of its U.S. operations nears, TikTok is hurrying to settle a privacy lawsuit in Illinois. The suit alleges that TikTok unlawfully recorded minors’ personal information, including face scans, both to be used for marketing purposes and to bolster further AI development in China. According to the Wall Street Journal, face geometry scans and voiceprints are collected by TikTok and used to recommend content based on users’ age, race, and physical attractiveness.
Sandvine ends Belarus services over human rights abuses
California tech company Sandvine, Inc. has terminated its internet services with the Belarus National Traffic Exchange Center because, it says, the local government used its products in a violation of human rights. The company provides technology used to filter and manage internet networks. When the highly-contested re-election of Belarus President Alexander Lukashenko occurred in August, the country’s National Traffic Exchange Center used Sandvine’s equipment to block Belarus citizens from accessing social media sites, messaging apps, and international news sites for several days. Sandvine said its equipment will not cease working in Belarus, but the company will no longer update or support those pieces. Read more on Bloomberg.
Facebook whistleblower exposes negligence and sluggishness
When Facebook data scientist Sophie Zhang was fired, she turned down a $64,000 non-disparagement package so that she could speak out against the slow movement, poor organization, and negligent behavior among the top brass at the social giant which has allowed multiple foreign governments to use the platform for criminal subterfuge such as undermining elections. “I have personally made decisions that affected national presidents without oversight, and taken action to enforce against so many prominent politicians globally that I’ve lost count,” Zhang wrote in the 6,600-word memo. Read the full story on BuzzFeed.
This week’s ‘must-read’ on The Avast Blog
The Covid-19 pandemic is forcing companies and designers to rethink the open-plan approach — and maybe rethink the idea of the “office” altogether. Are cubicles soon bound to make a comeback?
*** This is a Security Bloggers Network syndicated blog from Blog | Avast EN authored by Avast Blog. Read the original post at: https://blog.avast.com/the-good-bad-of-face-mask-recognition-avast