Community Technology and Telecommunications CommissionJuly 10, 2024

Item 2: AI and Surveillance in Public Places Presentation — original pdf

Thumbnail of the first page of the PDF
Page 1 of 33 pages

Being Watched Embedding Ethics in Public Cameras Building technical, legal, and social approaches to maximize the trusted use of public camera-generated video data Technology & Information Policy Institute Dr. Sharon Strover, Leo Cao University of Texas at Austin September 13, 2023 Community Technology and Telecommunications Commission, Austin TX & Good Systems Agenda Project Overview Research Approaches Major Issues & Questions Preliminary Conclusions Next Steps A multi -component research approach Understand conceptions of privacy among citizens, local journalists, and city officials Identify open records law at State level and relevant data management issues Build and test differential access models Design smart city literacy training and initiate citizen engagement for traffic control Research Approaches  Rely on experts, city officials, ordinary people, and city units in order to understand variations in informational norms.  Delphi survey  interviews  case studies (e.g., Austin Public Library)  focus groups  public survey (fall 2023)  workshops  Toward Differential Access & Civic Policies Smart City Technologies Austin Fire Department ShotSpotter (not in use in Austin, as of 2021) H.A.L.O. cameras CCTV cameras traffic cameras red light cameras Austin Transportation Department Privacy, Contexts, and Publicness  contextual integrity approach to examine norms& critical social values  actors: subjects, senders, receivers  types of information  transmission principles– consent, reciprocity, etc.  contexts  e.g., How do you think about camera-mounted drones used by fire departments when fighting fires?  s ubje ct: drone s , ca me ra s  s e nde r: fire de pa rtme nt  re ce ive r: re s ide nts , city units  type s : ima ge s of pe ople , buildings , e tc.  tra ns mis s ion: purpos e s , be ne fits , goa ls Research Methods  qualitative  illuminate meanings, values & norms associated with public technology + privacy  seven focus groups: identify benefits and concerns  seniors; students; EFF; civic activists; tech workers; tech executives; library users  in-depth case study: Austin Public Library  tour of camera system infrastructure,  group & individual interviews w/ IT, Security, Frontline staff  interviews with other City units:  Fire Department; Mobility (transportation); Chief Privacy Officer; Innovation Officer Core Findings: Focus Groups benefits & concerns  security aspects of surveillance tech  younger people assume no privacy anyway; make sure tech is used for "good" purposes  complications with personal uses of cameras in quasi-public spaces (Ring networks)  perception of public cameras influenced by popular culture – unaware of other uses  tech as not just a shortcut to dealing with bigger, complex problems... attributes of data  uncertain exchange: want transparency (data use, control) but question where data ‘go’  data governance needed at the local level  data & algorithmic literacy Core Findings: Focus Groups contexts  uncertainty of spatial qualities: What is public space anymore?  shared spaces such as apartment hallways  Ring cameras and other private systems see into public space  seeing the outside street from one's living spaces  drones complicate issue of publicness  real scenarios where tech is countered: contextual significance  no one-size-fits-all approach  aware of trade offs between privacy and security  awareness of Austin specifically as an increasingly tech city actors  people's own identity/background figures into how they think about risks  e.g., undocumented status, ICE investigations, traffic accidents Why the library? Cameras in public libraries  450 cameras, half in one downtown building  security & cameras "hiding in plain sight"?  library as trusted, open, welcoming, and safe site APL Website Privacy Statement The Austin Public Library is committed to the protection of all Library customers’ rights to privacy in the use of Library resources and discloses customer information to the customer only. Records of customer transactions are kept only as long as is required to collect administrative statistics and then are erased. Library records will only be disclosed under court order, subpoena, or warrant as outlined in the state statute, Texas Government Code, Section 552.124 and the surveillance provisions included in The USA PATRIOT ACT (Public Law 107-56). April 2003 Also: Library Use Rules 1. Purpose. The Austin Public Library is supported by the taxes of the people of Austin who expect each of our facilities to be clean, comfortable, and safe. The library is intended for reading, studying, writing and listening to written or electronically transmitted materials, and attending library or community sponsored programs and meetings. To this end, the library has established these rules to protect the rights and safety of library customers, staff and volunteers, and to help preserve and protect the library’s materials, equipment, facilities, and grounds … Research at the Library  observation & interviews  IT staff  Security staff  Frontline staff  Focus group with library users  7 additional focus groups investigating privacy norms with city technologies library IT server closet Research at the Library exhibit space with cameras and controlled entry with monitoring capabilities Research at the Library "[Removing the gates] goes along the direction of where libraries are more open, trusting . Let’s just trust everyone’s going to do the right thing vs. everyone’s evil trying to do bad things.” exhibit space with cameras and controlled entry with monitoring capabilities Research at the Library laptop checkout kiosk with camera Research at the Library Library as open, and safe space “We are here primarily to make sure that everyone here gets a good visit , that they feel safe when they come here. We do welcome all walks of life. So, we want to make sure everybody's treated fairly . And our biggest way of doing that is to be patrolling, be visible.” Research at the Library Tradeoff between security and discomfort “It seems like every day we're trying to balance between being welcoming and then ensuring safety and security. We're trying to find that middle ground, but as librarians we also feel like we need to be open for everybody and we want to welcome everybody but that goes against people. […] It will make them uncomfortable.” "The library seems to be mostly benign with how it uses its cameras. I think their making them more obvious, would create too much of an oppressive ambiance. People come to the library to kind of chill out and relax and be in a safe public space . Feeling like somebody's watching them, even if they are, it might be like too much of like a monkey on the back .” -library user Research at the Library Property theft “[…] the thing that stops them the most is the fact that it's a small space and we say hello to them […]. I don't think anybody's caring about the fact that there's cameras in there. It's more of the presence of us being in there that stops things from happening . It's also a strategy that we employ, […] greeting people so that they know 'I'm looking at you, I see you, I'm watching you.'” “[…] I like interacting with people a lot, and it has kind of like a dual thing of like making people feel seen and reminding them that they're like in a public space […], if that makes sense.” Research at the Library Blurred lines between public and private spaces “It's also unclear, at least to me where our ‘property’ ends . Because the Special Events Center downstairs is technically a city space and not the library space, so it isn't clear where these boundaries are for where the cameras shouldn't be covering or where securities jurisdiction comes up against.”-staff "I feel like I've just kind of personally, like, I justtake it for granted or I've grown up with the expectation that, oh yes, like every detail is like sitting on a server somewhere something like yeah, [I] kind of forget about it.” -library patron, on existence of camera data from library Research at the Library Cameras as reactive rather than proactive “We do our best internally to make sure that we have descriptions that can be shared and are recognizable because oftentimes we don't see those photos till much later whenever security gets around putting them onto the incident report or if they are added at all. So, I don't think we really depend on photos until it’s a repeating incident.” “Cameras aren’t really that much of a deterrent to people who are gonna act out or steal or anything like that. It might be a false sense of security for us in a way. I always feel like well, at least we would be able to go back and identify the person, hopefully if something happens or if they come back again, but it doesn't necessarily stop anything from happening .” Research at the Library Cameras as reactive rather than proactive “[…] if someone comes back enough that they have photos, oftentimes they're just recognizable to staff on site. So eventually the photos don't matter anymore . […] Like 'this person's back'. This is what they look like because sometimes finding the relevant incident report takes much more time than describing what this person looks like.” “[…] just like a lot of people have already said cameras are more reactive than proactive . I think it's more necessary to focus on how to end a situation in that moment as safely as possible rather than how to go back and see what happened.” Cameras in public libraries Wearable communication devices for quick response …. staff use a simple communication device worn around the neck. This radio was used to alert, coordinate, “rally the troops ,” and to keep track of one another. It offers a way to knit workers together and a less intrusive way to know who was doing what where, built on the consent and knowledge of those wearing the devices. Major issues & discussion  cameras as reactive, not proactive  straddling inclusivity and safety  library surveillance as proxy solution for bigger issues (e.g., housing, identification) Major issues & discussion  cameras as reactive, not proactive  straddling inclusivity and safety  library surveillance as proxy solution for bigger issues (e.g., housing, identification) Select References  Crawford, K. (2016). Big urban data and shrinking civic space: The statistical city meets the simulated city. In Barney, D. et al. (Ed.), The participatory condition in the digital age (pp. 169-226). Unive rs ity of Minne s ota P re s s .  Ma tte rn, S . (2021). A city is not a computer: Other urban intelligences. P rince ton Unive rs ity P re s s .  Kline nbe rg, E. (2018). Palaces for the people: How social infrastructure can help fight inequality, polarization, and the decline of civic life. Crown  Gre e ne , D. (2021). The promise of access: Technology, inequality and the political economy of hope. MIT P re s s .  S cott, J . C. (1998). Seeing like a state: How certain schemes to improve the human condition have failed. Ya le Unive rs ity P re s s .