Oakland, California On Its Way To Becoming The Third US City To Ban Facial Recognition Tech

from the beat-it,-face-jockeys dept

For the third time in two months, a US city has banned the use of facial recognition tech by local government agencies.

San Francisco started this movement (oh god please let it be a movement) back in May, booting the tech out of the city before local agencies had even gotten a chance to fool around with it. Earlier this month, Somerville, Massachusetts took home the silver in the anti-surveillance-state games, enacting a local ban on facial recognition tech.

Oakland, California has become the third city in the nation move forward with a facial recognition tech ban, as KPIX reports:

The city of Oakland on Tuesday took a first step towards banning the use of facial recognition technology by city employees, including the Oakland Police Department.

The City Council voted unanimously to approve the ordinance, which would prohibit police from both acquiring the software or using it at all, including if used by other police agencies.

This won't become official until September, when the council will reconvene for a second -- and final -- vote on the ordinance. It seems unlikely a majority of councilmembers will switch positions in the next 60 days, so local agencies with an eye on acquiring this tech should probably put those dreams to bed.

This ban appears to be slightly broader than those enacted elsewhere. It forbids government agencies from using the tech even when partnering with non-city agencies that might have this tech in their possession.

And a little homework goes a long way. The city council's president wrote a report on all the downsides of facial recognition tech, including its lack of accuracy and its tendency to err more often when attempting to identify minorities.

This ban will cramp the Oakland PD's style a bit. It already uses facial recognition tech via a database owned by the San Mateo County Sheriff. It argued that it should still be allowed to do this, but the council disagreed, voting for a ban that locks them out of the Sheriff's database. Hopefully, that language will be retained when it comes up for a final vote in September.

Filed Under: bans, facial recognition, oakland, privacy


Reader Comments

Subscribe: RSS

View by: Time | Thread


  • icon
    John Snape (profile), 22 Jul 2019 @ 3:34pm

    Now we just need to ban

    If we could now ban Stingrays and the TSA, we'd be closer to normalcy in this country.

    reply to this | link to this | view in chronology ]

  • identicon
    bobob, 22 Jul 2019 @ 10:47pm

    Three cities doth not a country make, except perhaps in Liechtenstein. I'm rather dubious that banning facial recognition will become a trend. If anyone remembers, DNA evidence was only supposed to exclude people as suspects and so it will go with facial recognition, unfortunately.

    reply to this | link to this | view in chronology ]

    • identicon
      Anonymous Coward, 23 Jul 2019 @ 6:42am

      Re:

      Thank you Mr Soothsayer,
      Those of us at Facial Recog International (FRI) who have stock options want to thank you for your support. Please continue with your belief in our crusade.

      I must point one thing however, and that is DNA is not the same as a picture. One could do a bit for bit comparison between two instances of pictures that seem to be the same in order to determine whether they are actually contain the same data, this cannot be said about DNA. As you pointed out, DNA can be used to exclude possible matches but fails when asked to confirm it is the same data. So, your conclusion is not necessarily correct.

      reply to this | link to this | view in chronology ]

      • identicon
        bobob, 23 Jul 2019 @ 10:54am

        Re: Re:

        If you work for a facial recognition company, you should know better than that. If you compare two different photos, you obviously cannot have the same data, bit by bit, so your use of "same data" is highly misleading, since the only possible way to refer to data from different images is to assume that you've extracted what is important to matching two images, which is exactly the problem. The only way a bitwise comparison of two images will be the same is if you are comparing an image to an exact bitwise copy of itself.

        Further,ore, you can read the following and the references therein to see how to manipulate images such that two images that appear the same will be determined to be different by AI software:

        https://arxiv.org/abs/1607.02533

        If that is possible, it is just as possible to take images of two different people which might look similar and manipulate them to cause AI to determine the images are of the same person. It's not like it's unknown for forensics labs to get the results they were predisposed to getting when the evidence would not support their conclusions.

        reply to this | link to this | view in chronology ]

  • This comment has been flagged by the community. Click here to show it
    icon
    Robin Ahmed (profile), 23 Jul 2019 @ 7:00am

    need help

    Does anyone talk this list for? best sewing machine for beginners https://www.bestsewingguide.com/best-sewing-machines-for-beginners/

    reply to this | link to this | view in chronology ]


Add Your Comment

Have a Techdirt Account? Sign in now. Want one? Register here



Subscribe to the Techdirt Daily newsletter




Comment Options:

  • Use markdown. Use plain text.
  • Remember name/email/url (set a cookie)

Close

Add A Reply

Have a Techdirt Account? Sign in now. Want one? Register here



Subscribe to the Techdirt Daily newsletter




Comment Options:

  • Use markdown. Use plain text.
  • Remember name/email/url (set a cookie)

Follow Techdirt
Techdirt Gear
Shop Now: Techdirt Logo Gear
Advertisement
Report this ad  |  Hide Techdirt ads
Essential Reading
Techdirt Deals
Report this ad  |  Hide Techdirt ads
Techdirt Insider Chat
Advertisement
Report this ad  |  Hide Techdirt ads
Recent Stories
Advertisement
Report this ad  |  Hide Techdirt ads

Close

Email This

This feature is only available to registered users. Register or sign in to use it.