San Francisco, Aug 22 : Tech giant Google’s artificial intelligence (AI) has been reported to have identified parents’ accounts for potential abuse in relation to naked images of their sick children.The father explained that when he used his Android smartphone to capture pictures of an infection in his child’s groin, the company flagged the photos as child sexual assault material (CSAM) and cited the NYT, The Verge reported.
The company shut his accounts , and he filed a complaint with the National Center for Missing and Exploited Children (NCMEC) and resulted in an investigation by the police.
It highlights the difficulties of trying to discern the difference between an abused image and a harmless photograph once it’s part of the user’s digital library regardless of whether it is it is on their device at home or in cloud storage.
The incident happened in February 2021 when certain doctor’s offices were closed due to the Covid-19 epidemic.
According to the article, Mark (whose last name was not disclosed) was able to see swelling in his child’s genital area and on the instruction of an aide, sent pictures of the problem prior to the start of an appointment via video.
The doctor ended up prescribing antibiotics to treat the infection.
Mark received a notice from Google two days after taking the photos, notifying him that his accounts were shut down for “harmful material” which is “a serious infraction of Google’s policies” and could constitute illegal.”
Like many other internet companies, such as Facebook, Twitter, and Reddit, Google has used hash-matching with Microsoft’s PhotoDNA to scan uploaded photos to identify the matches with well-known CSAM.
In 2012 it resulted in the detention of an individual who was an registered sex offender who made use of Gmail to send pictures of young girls.
vc/kvd
