<html>
  <head>

    <meta http-equiv="content-type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <div class="container content-width3" style="--font-size:20px;">
      <div class="header reader-header reader-show-element"> <font
          size="-2"><a class="domain reader-domain"
href="https://theintercept.com/2020/01/27/surveillance-cctv-smart-camera-networks/">https://theintercept.com/2020/01/27/surveillance-cctv-smart-camera-networks/</a></font>
        <h1 class="reader-title">The Rise of Smart Camera Networks, and
          Why We Should Ban Them</h1>
        <div class="credits reader-credits">Michael Kwet - January 27,
          2020<br>
        </div>
      </div>
      <hr>
      <div class="content">
        <div class="moz-reader-content line-height4 reader-show-element">
          <div id="readability-page-1" class="page">
            <div data-reactid="210">
              <div data-reactid="211">
                <p><u>There’s widespread concern</u> that video cameras
                  will use facial recognition software to track our
                  every public move. Far less remarked upon — but every
                  bit as alarming — is the exponential expansion of
                  “smart” video surveillance networks.</p>
                <p>Private businesses and homes are starting to plug
                  their cameras into police networks, and rapid advances
                  in artificial intelligence are investing
                  closed-circuit television, or CCTV, networks with the
                  power for total public surveillance. In the
                  not-so-distant future, police forces, stores, and city
                  administrators hope to film your every move — and
                  interpret it using video analytics.</p>
                <p>The rise of all-seeing smart camera networks is an
                  alarming development that threatens civil rights and
                  liberties throughout the world. Law enforcement
                  agencies have a long history of using surveillance
                  against marginalized communities, and studies show
                  surveillance chills freedom of expression — ill
                  effects that could spread as camera networks grow
                  larger and more sophisticated.</p>
                <p>To understand the situation we’re facing, we have to
                  understand the rise of the video surveillance
                  industrial complex — its history, its power players,
                  and its future trajectory. It begins with the
                  proliferation of cameras for police and security, and
                  ends with a powerful new industry imperative: complete
                  visual surveillance of public space.</p>
                <h3>Video Management Systems and Plug-in Surveillance
                  Networks</h3>
                <p>In their first decades of existence, CCTV cameras
                  were low-resolution analog devices that recorded onto
                  tapes. Businesses or city authorities deployed them to
                  film a small area of interest. Few cameras were placed
                  in pubic, and the power to track people was limited:
                  If police wanted to pursue a person of interest, they
                  had to spend hours collecting footage by foot from
                  nearby locations.</p>
                <p>In the late 1990s, video surveillance became more
                  advanced. A company called Axis Communications
                  invented the first <a
                    href="https://www.axis.com/newsroom/article/first-network-camera">internet-enabled
                    surveillance camera</a>, which converted moving
                  images to digital data. New businesses like Milestone
                  Systems built Video Management Systems, or VMS, to
                  organize video information into databases. VMS
                  providers created new features like motion sensor
                  technology that alerted guards when a person was
                  caught on camera in a restricted area.</p>
                <p>As time marched on, video surveillance spread. On <a
href="https://www.amazon.com/Business-Magnetism-Mr-Lars-Thinggaard/dp/0615892140">one
                    account</a>, about 50 years ago, the United Kingdom
                  had somewhere north of 60 permanent CCTV cameras
                  installed nationwide. Today, the U.K. has over <a
                    href="https://www.bbc.com/news/uk-30978995">6
                    million</a> such devices, while the U.S. has <a
href="https://technology.ihs.com/583114/north-american-security-camera-installed-base-to-reach-62-million-in-2016">tens
                    of millions</a>. According to marketing firm IHS
                  Markit, <a
href="https://www.wsj.com/articles/a-billion-surveillance-cameras-forecast-to-be-watching-within-two-years-11575565402">1
                    billion cameras</a> will be watching the world by
                  the end of 2021, with the United States rivaling
                  China’s <a
href="https://technology.ihs.com/619515/the-us-has-a-security-camera-penetration-rate-rivalling-chinas">per
                    person camera penetration rate</a>. Police can now
                  track people across multiple cameras from a
                  command-and-control center, desktop, or smartphone.</p>
                <p>While it is possible to link thousands of cameras in
                  a VMS, it is also expensive. To increase the amount of
                  CCTVs available, cities recently came up with a clever
                  hack: encouraging businesses and residents to place
                  privately owned cameras on their police network — what
                  I call “plug-in surveillance networks.”</p>
              </div>
              <div data-reactid="212">
                <div data-reactid="213">
                  <p><img
src="https://theintercept.imgix.net/wp-uploads/sites/1/2020/01/h_15235476-edit-1579897222.jpg?auto=compress%2Cformat&q=90&w=1024&h=682"
                      alt="Video from surveillance cameras around the
                      city is displayed at the Real Time Crime Center
                      the viewing space for Project Green Light, at the
                      Police Department's headquarters in downtown
                      Detroit, June 14, 2019. In recent weeks, a public
                      outcry has erupted over the facial recognition
                      program employed in conjunction with the network
                      of cameras. (Brittany Greeson/The New York Times)"></p>
                  <p class="caption">Video from surveillance cameras
                    around the city is displayed at the Real-Time Crime
                    Center, the viewing space for Project Green Light,
                    at the police department headquarters in Detroit on
                    June 14, 2019.</p>
                  <p class="caption">
                    Photo: Brittany Greeson/The New York Times via Redux</p>
                </div>
              </div>
              <div data-reactid="214">
                <p>By pooling city-owned cameras with privately owned
                  cameras, policing experts say an agency in a typical
                  large city may amass <a
                    href="https://www.rand.org/pubs/research_reports/RR2619.html">hundreds
                    of thousands</a> of video feeds in just a few years.</p>
                <p>Detroit has popularized plug-in surveillance networks
                  through its controversial Project Green Light program.
                  With Project Green Light, businesses can purchase CCTV
                  cameras and connect them to police headquarters. They
                  can also place a bright green light next to the
                  cameras to indicate they are part of the police
                  network. The project claims to deter crime by
                  signaling to residents: The police are watching you.</p>
                <p>Detroit is not alone. <a
                    href="https://www.chicago.gov/city/en/depts/oem/provdrs/tech.html">Chicago</a>,
                  <a href="https://www.safecamnola.com/">New Orleans</a>,
                  <a
href="https://www.cityandstateny.com/articles/policy/technology/how-new-york-city-is-watching-you.html">New
                    York</a>, and <a
                    href="https://atlantapolicefoundation.org/programs/operation-shield/">Atlanta</a>
                  have also deployed plug-in surveillance networks. In
                  these cities, private businesses and/or homes provide
                  feeds that are integrated into crime centers so that
                  police can access live streams and recorded footage.
                  The police department in New Haven, Connecticut, told
                  me they are looking into plug-in surveillance, and
                  others are likely considering it.</p>
                <p>The number of cameras on police networks now range
                  from tens of thousands (Chicago) to several hundred
                  (New Orleans). With so many cameras in place, and only
                  a small team of officers to watch them, law
                  enforcement agencies face a new challenge: How do you
                  make sense of all that footage?</p>
                <p>The answer is video analytics.</p>
                <h3>Video Analytics Takes Off</h3>
                <p>Around 2006, a young Israeli woman was recording
                  family videos every weekend, but as a student and
                  parent, she didn’t have time to watch them. A computer
                  scientist at her university, Professor Shmuel Peleg,
                  told me he tried to create a solution for her: He
                  would take a long video and condense the interesting
                  activity into a short video clip.</p>
                <p>His solution failed: It only worked on stationary
                  cameras, and the student’s video camera was moving
                  when she filmed her family.</p>
                <p>Peleg soon found another use case in the surveillance
                  industry, which relies on stationary cameras. His
                  solution became BriefCam, a video analytics firm that
                  can summarize video footage from a scene across time
                  so that investigators can view all relevant footage in
                  a short space of time.</p>
              </div>
              <div data-reactid="217">
                <p>Using a feature called Video Synopsis, BriefCam
                  overlays footage of events happening at different
                  times<a
                    href="https://www.youtube.com/watch?v=fISfDd35sXU">
                    as if they are appearing simultaneously. </a>For
                  example, if several people walked past a camera at
                  12:30 p.m., 12:40 p.m., and 12:50 p.m., BriefCam will
                  aggregate their images into a single scene.
                  Investigators can view all footage of interest from a
                  given day in minutes instead of hours.</p>
                <p>Thanks to rapid advances in artificial intelligence,
                  summarization is just one feature in BriefCam’s
                  product line and the rapidly expanding <a
                    href="https://www.aclu.org/report/dawn-robot-surveillance">video
                    analytics industry</a>.</p>
                <p>Behavior recognition includes video analytics
                  capabilities like <a
                    href="https://www.youtube.com/watch?v=QcCjmWwEUgg">fight
                    detection</a>, emotion recognition, fall detection,
                  loitering, dog walking, jaywalking, toll fare evasion,
                  and even <a
href="https://www.securitysales.com/news/milestone-disruption-innovation-mips/slideshow/2/">lie
                    detection</a>.</p>
                <p>Object recognition can recognize faces, animals,
                  cars, weapons, fires, and other things, as well as
                  human characteristics like gender, age, and hair
                  color.</p>
                <p>Anomalous or unusual behavior detection works by
                  recording a fixed area for a period of time — say, 30
                  days — and determining “normal” behavior for that
                  scene. If the camera sees something unusual — say, a
                  person running down a street at 3:00 a.m. — it will
                  flag the incident for attention.</p>
                <p>Video analytics systems can analyze and search across
                  real-time streams or recorded footage. They can also
                  isolate individuals or objects as they traverse a
                  smart camera network.</p>
                <p>Chicago; New Orleans; Detroit; Springfield,
                  Massachusetts; and Hartford, Connecticut, are some of
                  the cities currently using BriefCam for policing.</p>
                <h3>To Search and Surveil</h3>
                <p>With city spaces blanketed in cameras, and video
                  analytics to make sense of them, law enforcement
                  agencies gain the capacity to record and analyze
                  everything, all the time. This provides authorities
                  the power to index and search a vast database of
                  objects, behaviors, and anomalous activity.</p>
                <p>In Connecticut, police have used video analytics to
                  identify or monitor known or suspected drug dealers.
                  Sergeant Johnmichael O’Hare, former Director of the
                  Hartford Real-Time Crime Center, recently <a
                    href="https://www.youtube.com/watch?v=lDC3EVAMvYI">demonstrated</a>
                  how BriefCam helped Hartford police reveal “where
                  people go the most” in the space of 24 hours by
                  viewing footage condensed and summarized in just nine
                  minutes. Using a feature called “pathways,” he
                  discovered hundreds of people visiting just two houses
                  on the street and secured a search warrant to verify
                  that they were drug houses.</p>
                <p>Video analytics startup Voxel51 is also adding more
                  sophisticated searching to the mix. Co-founded by
                  Jason Corso, a professor of electrical engineering and
                  computer science at the University of Michigan, the
                  company offers a platform for video processing and
                  understanding.</p>
                <p>Corso told me his company hopes to offer the first
                  system where people can “search based on semantic
                  content about their data, such as, ‘I want to find all
                  the video clips that have more than 3-way
                  intersections … with at least 20 vehicles during
                  daylight.’” Voxel51 “tries to make that possible” by
                  taking video footage and “turning it into structured
                  searchable data across different types of platforms.”</p>
                <p>Unlike BriefCam, which analyzes video using nothing
                  but its own software, Voxel51 offers an open platform
                  which allows third parties to add their own analytics
                  models. If the platform succeeds, it will supercharge
                  the ability to search and surveil public spaces.</p>
                <p>Corso told me his company is working on a <a
href="https://techcrunch.com/2019/08/08/voxel51-raises-2-million-for-its-video-native-identification-of-people-cars-and-more/">pilot
                    project</a> with the Baltimore police for their
                  CitiWatch surveillance program and plans to trial the
                  software with the Houston Police Department.</p>
                <p>As cities start deploying a wide range of monitoring
                  devices from the so-called internet of things,
                  researchers are also developing a technique known as <a
href="https://www.rand.org/pubs/research_reports/RR2619.html">v</a><a
                    href="https://www.rand.org/pubs/research_reports/RR2619.html">ideo
                  </a><a
                    href="https://www.rand.org/pubs/research_reports/RR2619.html">a</a><a
href="https://www.rand.org/pubs/research_reports/RR2619.html">nalytics
                    and </a><a
                    href="https://www.rand.org/pubs/research_reports/RR2619.html">s</a><a
href="https://www.rand.org/pubs/research_reports/RR2619.html">ensor </a><a
href="https://www.rand.org/pubs/research_reports/RR2619.html">f</a><a
                    href="https://www.rand.org/pubs/research_reports/RR2619.html">usion</a>,
                  or VA/SF, for police intelligence. With VA/SF,
                  multiple streams from sensors are combined with video
                  analytics to reduce uncertainties and make inferences
                  about complex situations. As one example, Peleg told
                  me BriefCam is developing in-camera audio analytics
                  that uses microphones to discern actions that may
                  confuse AI systems, such as whether people are
                  fighting or dancing.</p>
                <p>VMSs also offer smart integration across
                  technologies. Former New Haven Chief of Police Anthony
                  Campbell told me how <a
                    href="https://www.youtube.com/watch?time_continue=7&v=JJSNQoIvvkY">ShotSpotters</a>,
                  controversial devices that listen for gunshots,
                  integrate with specialized software so when a gun is
                  fired, nearby swivel cameras instantly <a
href="https://sixtechsys.com/the-hawkeye-effect-recognized-again-for-its-innovative-software">alter
                    their direction</a> to the location of the weapons
                  discharge.</p>
                <p>Officers can also use software to lock building doors
                  from a control center, and companies are developing
                  analytics to alert security if <a
                    href="https://www.eblockwatch.co.za/posts/watch-the-future">one
                    car is being followed by another</a>.</p>
                <h3>Toward a “Minority Report” World</h3>
                <p>Video analytics captures a wide variety of data about
                  the areas covered by smart camera networks. Not
                  surprisingly, the information captured is now being
                  proposed for predictive policing: the use of data to
                  predict and police crime before it happens.</p>
                <p>In 2002, the dystopian film “Minority Report<i>”</i>
                  depicted a society using “pre-crime” analytics for
                  police to intervene in lawbreaking before it occurs.
                  In the end, the officers in charge tried to manipulate
                  the system for their own interests.</p>
                <p>A real-world version of “Minority Report” is <a
href="https://www.vice.com/en_us/article/7xmmvy/why-does-hartford-have-so-many-cameras-precrime">emerging</a>
                  through real-time crime centers used to analyze crime
                  patterns for police. In these centers, law enforcement
                  agencies ingest information from sources like social
                  media networks, data brokers, public databases,
                  criminal records, and ShotSpotters. Weather data is
                  even included for its impact on crime (because “bad
                  guys don’t like to get wet”).</p>
                <p>In a <a
href="https://www.accenture.com/_acnmedia/pdf-94/accenture-value-data-seeing-what-matters.pdf">2018
                    document</a>, the data storage firm Western Digital
                  and the consultancy Accenture predicted mass smart
                  camera networks would be deployed “across three tiers
                  of maturity.” This multi-stage adoption, they
                  contended, would “allow society” to gradually abandon
                  “concerns about privacy” and instead “accept and
                  advocate” for mass police and government surveillance
                  in the interest of “public safety.”</p>
                <p>Tier 1 encompasses the present where police use CCTV
                  networks to investigate crimes after-the-fact.</p>
                <p>By 2025, society will reach Tier 2 as municipalities
                  transform into “smart” cities, the document said.
                  Businesses and public institutions, like schools and
                  hospitals, will plug camera feeds into government and
                  law enforcement agencies to inform centralized,
                  AI-enabled analytics systems.</p>
                <p>Tier 3, the most predictive-oriented surveillance
                  system, will arrive by 2035. Some residents will
                  voluntarily donate their camera feeds, while others
                  will be “encouraged to do so by tax-break incentives
                  or nominal compensation.” A “public safety ecosystem”
                  will centralize data “pulled from disparate databases
                  such as social media, driver’s licenses, police
                  databases, and dark data.” An AI-enabled analytics
                  unit will let police assess “anomalies in real time
                  and interrupt a crime before it is committed.”</p>
                <p>That is to say, to catch pre-crime.</p>
                <h3>Rise of the Video Surveillance Industrial Complex</h3>
                <p>While CCTV surveillance began as a simple tool for
                  criminal justice, it has grown into a
                  multibillion-dollar industry that covers multiple
                  industry verticals. From policing and smart cities to
                  schools, health care facilities, and retail, society
                  is moving toward near-complete visual surveillance of
                  commercial and urban spaces.</p>
                <p>Denmark-based Milestone Systems, a top VMS provider
                  with half its revenues in the U.S., had less than 10
                  employees in 1999. Today they are a major corporation
                  that claims offices in over 20 countries.</p>
                <p>Axis Communications used to be a network printer
                  outfit. They have since become a leading camera
                  provider pushing over $1 billion in sales per year.</p>
                <p>BriefCam began as a university project. Now it is
                  among the world’s top video analytics providers, with
                  clients, it says, spanning <a
href="https://www.briefcam.com/company/press-releases/briefcam-achieves-exponential-growth-demand-surges-award-winning-video-synopsis-deep-learning-technology/">over
                    40 countries</a>.</p>
                <p>Over the past six years, Canon purchased all three,
                  giving the imaging conglomerate ownership of industry
                  giants in video management software, CCTV cameras, and
                  video analytics. Motorola recently acquired a top VMS
                  provider, Avigilon, for $1 billion. In turn, Avigilon
                  and other large firms have purchased their own
                  companies.</p>
              </div>
              <blockquote data-reactid="218"><span data-reactid="219"></span>
                <p>The public is paying for their own high-tech
                  surveillance three times over.</p>
              </blockquote>
              <div data-reactid="221">
                <p>Familiar big tech giants are also in on the action.
                  Lieutenant Patrick O’Donnell of the Chicago police
                  force told me his department is working on a
                  non-disclosure agreement with Google for a video
                  analytics pilot project to detect people reacting to
                  gunfire, and if they are in the prone position, so the
                  police can receive real-time alerts. (Google did not
                  respond to a request for comment.)</p>
                <p>Video monitoring networks inevitably entangle and
                  implicate a whole ecosystem of vendors, some of whom
                  have offered, or may yet offer, services specifically
                  targeted at such systems. Microsoft, Amazon, IBM,
                  Comcast, Verizon, and Cisco are among those enabling
                  the networks with technologies like cloud services,
                  broadband connectivity, or video surveillance
                  software.</p>
                <p>In the public sector, the National Institute of
                  Standards and Technology is <a
href="https://www.nist.gov/news-events/news/2017/06/nist-awards-385-million-accelerate-public-safety-communications">funding</a>
                  “public analytics” and communications networks like
                  the First Responder Network Authority, or <a
                    href="https://theintercept.com/2018/07/29/firstnet-att-surveillance/">FirstNet,</a>
                  for real-time video and other surveillance
                  technologies. FirstNet will cost $46.5 billion, and is
                  being built by AT&T.</p>
                <p>Voxel51 is another NIST-backed venture. The public is
                  thus paying for their own high-tech surveillance three
                  times over: first, through taxes for university
                  research; second, through grant money for the
                  formation of a for-profit startup (Voxel51); and
                  third, through the purchase of Voxel51’s services by
                  city police departments using public funds.</p>
                <p>With the private and public sector looking to expand
                  the presence of cameras, video surveillance has become
                  a new cash cow. As Corso put it, “there will be
                  something like 45 billion cameras in the world within
                  a few decades. That’s a lot of (video) pixels. For the
                  most part, most of those pixels go unused.” Corso’s
                  estimate mirrors a <a
href="https://www.fastcompany.com/40450867/in-less-than-five-years-45-billion-cameras-will-be-watching-us">2017
                    forecast</a> from New York venture capital firm LDV,
                  which believes smartphones will evolve to have even
                  more cameras than they do today, contributing to the
                  growth.</p>
                <p>Companies that began with markets for police and
                  security are now diversifying their offerings to the
                  commercial sector. BriefCam, Milestone, and Axis
                  advertise the use of video analytics for retailers,
                  where they can monitor foot traffic, queue length,
                  shopping patterns, floor layouts, and conduct <a
href="https://www.briefcam.com/solutions/consumer-behavior-and-experience/">A/B
                    testing</a>. Voxel51 has an option built for the
                  fashion industry and plans to expand across industry
                  verticals. <a href="https://motionloft.com/">Motionloft</a>
                  offers analytics for smart cities, retailers,
                  commercial real estate, and entertainment venues.
                  Other examples abound.</p>
                <p>Public and private sector actors are pressing for a
                  world full of smart video surveillance. Peleg, for
                  example, told me of a use case for smart cities: If
                  you drive into the city, you could “just park and go
                  home” without using a parking meter. The city would
                  send a bill to your house at the end of the month. “Of
                  course, you lose your privacy,” he added. “The
                  question is, do you really care about Big Brother
                  knows where you are, what you do, etc.? Some people
                  may not like it.”</p>
                <h3>How to Rein in Smart Surveillance</h3>
                <p>Those who do not like new forms of Big Brother
                  surveillance are presently fixated on facial
                  recognition. Yet they have largely ignored the shift
                  to smart camera networks — and the industrial complex
                  driving it.</p>
                <p>Thousands of cameras are now set to scrutinize our
                  every move, informing city authorities whether we are
                  walking, running, riding a bike, or doing anything
                  “suspicious.” With video analytics, artificial
                  intelligence is used to identify our  sex, age, and
                  type of clothes, and could potentially be used to
                  categorize us by race or religious attire.</p>
                <p>Such surveillance could have a severe chilling effect
                  on our freedom of expression and association. Is this
                  the world we want to live in?</p>
                <p>The capacity to track individuals across smart CCTV
                  networks can be used to target marginalized
                  communities. The detection of “loitering” or
                  “shoplifting” by cameras concentrated in poor
                  neighborhoods may deepen racial bias in policing
                  practices.</p>
                <p>This kind of racial discrimination is <a
href="https://www.vice.com/en_us/article/pa7nek/smart-cctv-networks-are-driving-an-ai-powered-apartheid-in-south-africa">already
                    happening</a> in South Africa, where “unusual
                  behavior detection” has been deployed by smart camera
                  networks for several years.</p>
                <p>In the United States, smart camera networks are just
                  emerging, and there is little information or
                  transparency about their use. Nevertheless, we know
                  surveillance has been used throughout history to
                  target oppressed groups. In recent years, the New York
                  Police Department secretly spied on Muslims, the FBI
                  used surveillance aircraft to monitor Black Lives
                  Matter protesters, and the U.S. Customs and Border
                  Protection began building a high-tech video
                  surveillance “<a
href="https://theintercept.com/2019/08/25/border-patrol-israel-elbit-surveillance">smart
                    border</a>” across the Tohono O’odham reservation in
                  Arizona.</p>
                <p>Law enforcement agencies claim smart camera networks
                  will reduce crime, but at what cost? If a camera could
                  be put in every room in every house, domestic violence
                  might go down. We could add automated “filters” that
                  only record when a loud noise is detected, or when
                  someone grabs a knife. Should police put smart cameras
                  inside every living room?</p>
                <p>The commercial sector is likewise rationalizing the
                  advance of surveillance capitalism into the physical
                  domain. Retailers, employers, and investors want to
                  put us all under smart video surveillance so they can
                  manage us with visual “intelligence.”</p>
                <p>When asked about privacy, several major police
                  departments told me they have the right to see and
                  record everything you do as soon as you leave your
                  home. Retailers, in turn, won’t even approach public
                  disclosure: They are keeping their video analytics
                  practices <a
href="https://www.aclu.org/blog/privacy-technology/surveillance-technologies/are-stores-you-shop-secretly-using-face">secret</a>.</p>
                <p>In the United States, there is generally no
                  “reasonable expectation” of privacy in public. The
                  Fourth Amendment encompasses the home and a few public
                  areas we “reasonably” expect to be private, such as a
                  phone booth. Almost everything else — our streets, our
                  stores, our schools — is fair game.</p>
                <p>Even if rules are updated to restrict the <em>use</em>
                  of video surveillance, we cannot guarantee those rules
                  will remain in place. With thousands of high-res
                  cameras networked together, a dystopian surveillance
                  state is a mouse click away. By installing cameras
                  everywhere, we are opening a Pandora’s box.</p>
                <p>To address the privacy threats of smart camera
                  networks, legislators should ban plug-in surveillance
                  networks and restrict the scope of networked CCTVs
                  beyond the premise of a single site. They should also
                  limit the density of camera and sensor coverage in
                  public. These measures would block the capacity to
                  track people across wide areas and prevent the
                  phenomenon of constantly being watched.</p>
                <p>The government should also ban video surveillance
                  analytics in publicly accessible spaces, perhaps with
                  exceptions for rare cases such as the detection of
                  bodies on train tracks. Such a ban would
                  disincentivize mass camera deployments because video
                  analytics is needed to analyze large volumes of
                  footage. Courts should urgently <a
href="https://digitalcommons.wcl.american.edu/cgi/viewcontent.cgi?referer=&httpsredir=1&article=1899&context=aulr">reconsider</a>
                  the scope of the Fourth Amendment and expand our <a
href="http://law.emory.edu/elj/_documents/volumes/66/3/levinson-waldman.pdf">right
                    to privacy in public</a><a
href="http://law.emory.edu/elj/_documents/volumes/66/3/levinson-waldman.pdf">.</a></p>
                <p>Police departments, vendors, and researchers need to
                  disclose and publicize their projects, and engage with
                  academics, journalists, and civil society.</p>
                <p>It is clear we have a crisis in the works. We need to
                  move beyond the limited conversation of facial
                  recognition and address the broader world of video
                  surveillance, before it is too late.</p>
              </div>
            </div>
          </div>
        </div>
      </div>
      <div> </div>
    </div>
    <div class="moz-signature">-- <br>
      Freedom Archives
      522 Valencia Street
      San Francisco, CA 94110
      415 863.9977
      <a class="moz-txt-link-freetext" href="https://freedomarchives.org/">https://freedomarchives.org/</a></div>
  </body>
</html>