"75 Million Criminal Law Enforcement Files" Database Added to Knightscope Facial Recognition
Knightscope, is adding a database of "over 75 million criminal law enforcement files" to its security robot's facial recognition but where are they getting it from and how will it work?
In this report, IPVM examines the announcement, including who the database supplier is, privacy concerns, and comparisons to other facial recognition companies, such as Clearview AI, and analyzes the potential implications and effectiveness of this offering.
See IPVM's previous coverage of Knightscope, including:
- Car/Security Robot Hit and Run
- Knightscope Announces a Non-Robot Intercom
- Knightscope Revenue up 2%, Net Loss up 126% in 2021
- Knightscope Goes Public, Plunges, Then Rockets, Then Falls
- Knightscope Going Public, Boasts "Main Street First, Wall Street Second"
Overview: "** ******* ******** *** *********** *****"
******************** * *** ********, ********* ******** Detection ********** (****) ** ********* *. The ******* ********* *** *** ******* as:
**** ******** *** ****** *********** ********** on ** **** *** * ***** open-source ******** **over ** ******* ******** *** *********** ***** to detect and alert users to a person previously convicted of a crime or possible threats on their property.
*********** *** *** ******* **** **** of "******** *** *********** *****" *** database ******** **, *** *** *** reply ** *** ******* *** ******* and *************.
Captis ******** ********
*********** ** *********** ************, ***** *********** ** ********* * criminal ******* ********, ****** ********* ** IPVM. ****** ****** ******* "************ **** over ** ******* ****** ******** *****". Captis's*** *** **-******* ***** ********** **** *** ******** *** *** convicted ******** *** *** *** ******* those ******** *** *** ********* *** those ********* *** ************. *** ******* did *** ******* *** ***** ****** number ** *********** ** *** ******** but *** ******* **** ****** ********* may **** ******** *****.
**** ****** ******' ******* ********** ****, ********** ** ** ******** plugin. ** ***** ****:
***** *** *********** ****** **** *** the ****** *********** ******** ******* **** look ******* ** ******** ******** ** their ********, *** **** ******* ******* was **** ** *** **** *** us ** ****** ******* ***** ******* were ******* ** ****** *************. *** risk, *********, ** **** ****** ***** be ************* ** ********* *********.
Facial *********** ********
** ****** *********** ******** ** *********** tries ** ***** ******* ****** **** images ** ****** ***** ****** *******. The ***** ********* * ***** **** ** ******** from * *********** / *** *********, ***** ******* ***** ****** ***** can ** *********:
******** *******: (*) *** *****'* ****** is ******* ********** ***, ** *** angle ** *** **** ** ****, (2) ****** **** ** ** ****** around, ********** *** *********** **** *** angle ** *** **** ** **** as **** *** (*) ***** ****** are ********** **** ******** ***** ******* light ******* (*** ***, *** **********, streetlights, ***.) **** ******* ***** ******* even *******.
Privacy ********
*** *** ** ****** *********** ********** is ************* *** ******* **** ********* privacy *** ***** ****** ********.
**** ****** ********** ****** ********** ******* and *****-*********** ** **** ******** *********, the **** *** *****—*** *** ************ for ***** ********—** **** *********** ** well,** ** **** **** **** ***** types ** ****** *********** *******.
Contrast ** ********* **
********* ** ** **** ***** *** its ************* ******* ********. *** *** York-based ******* *** ******* **** ** billion ****** **** ********, *********, *** Twitter,********* ** ******* **** ********* **** Computing.co.uk.
******** ** *********, ***** **** *** scraped ****** ** ******* *** *** facial *********** ********, ***********'* ******** ********* only ******** "********" ***** ******* **** publicly ********* *********** ** *****, ******, federal, *** ********** ********.
Practical ****** ***
** ****** ***********'* ******* ** **** little ******, ***** *** ******** ** doing ****** *********** **** * ***** and ***********'* ********* ***** ******** **** (of **** **** *** ******). *** combination ***** **** **** **** *** be ****** ****.
Sign ** ****** ******
***** ***********'* ****** ** ****** *******, we *** **** ** * **** of * ******* ***** **** ********* understand *** ***** *** ***** ** connecting * ******** ** ******** ***** to ***** ****** ***********. ****** ***********, by ******, ****** ******* ***** ******* the ********** ************* ******* ** ****** a ********* ** ***** *******. ********* like *********, ***** ************* *** **** violating *******, ******* ******** *** ********* power *** ****** *********** *****.
* **** *** ** *** *** of ***** ****** ** **** ****. Looks **** ****'** ****** ******** ** paper ***** ******** ******* ** **. Kind * **** *** ***** ** cryptocurrency *** ***'*
**** ******** *** ****** *********** ********** on ** ****
*** ** ** ***** **** ******, not ***** ***** ******
* *** ***** *** *** ***** of * ** ****** ****** ** a ********* ************** - ***** ***** **** ** alerted ** - *** *** **** a ****** * ******** ** ********* felons ***** **** *** ** ** anything ** ***?
*******: ** ***** ****** ******** ****** * ******* ***** ** ******** *** **** * hit ******* ** ***** ***, **** when ** *** ****** ****** *******, he *** **** ****** ******* * eight ***** *** * ****** ***** with *******.
**** **** *** ** ** **** that ******** ***?
*** **** ** *********** ***** ******* if *****’* * “***-********” *****. ** you’re ********** * ******, ****** ****** be *** *** ******** **** ** registered *** *********, ****** ********, *** missing/trafficked ********. ** ***’** ********** * port ** ***** (*.*. *******) ********** categories **** ** ********** *** ********** individuals ****** ** *****.
** * “****-********” *********** (*.*. ***** a ******* ** ******) ** ******** across *** ********** ***** ** *****.
****, ***** ***** ******** ** ******* at *** **** ********* ** **** reporting ******, *** **-**** *** **** of “****** *******” ***** ** **** cleared * **** **** *** ** Johnny ************ ****** *** **** *** didn’t ******* *** *********.
****** ****** ** *** *** ******** such ** ********** *** *********, ****** suspects, *** *******/********** ********.
****** ***** **** ** **** ** be *** *** * ** ***** 3 (********** *** *********), ** ********** is ***** ******
******'**** *** **-******* ***** ********** **** *** ******** *** ***convicted ******** and did not include those arrested but not convicted nor those convicted for misdemeanors.
****** ******** (****** ***** ****** *** escape ** ******/********* ********** - *.*. already ********* ******) *** *******/********** ******** are *** ******** ** *** ****** database.
****, ***** ***** ******** ** ******* at *** **** ********* ** **** reporting ******, *** **-**** *** **** of “****** *******” ***** ** **** cleared * **** **** *** ** Johnny ************ ****** *** **** *** didn’t ******* *** *********.
*** ***** *** (**** ******) ****** Johnson ** '*******'? **** ** *** survives *** ********* ******* ********* *** is ***** * ********* ***** - subject ** ********* ** *** ****** database.
"** ***’** ********** * **** ** entry (*.*. *******) ********** ********** **** as ********** *** ********** *********** ****** be *****." - ***** ** * common ************* **** *** ********** **** secretive ****** ** *** ********* **********. Having * ***** ******** ********** **** not ******** ******* **** *********. *********, there ** * **-****** ******** ***** may ******* ******** **** ****** *** main ** ********* ** *** ****/ DOB/ ******** ****** (*** ************* ******), that ******** ** ******* ** *** airlines (** *** *************) *** ***** is * ****** ******* ** "**********" into **** ********* ******** ********* **** the ***** ** ******** ******** ***** the ******* ***. ***** ** ** room *** * ****** **** *********** at ******** ** ***** ** ***** where * ******** ******* (**** * TSA ********** ** ** *********** *****) is ******** ** *) ***** *** travelers *** *) **** *** **** up *** *********** ** ****** ** need **.
**** ** ********, * ******* ******* violation ******* ** ********* *******. ** derives * *********** ** ***** (********** alarms ** * ******** ********) ***** on ********** *** **** ******* ********** (the ****** *** ******** ***, *** worse *** ******** **; **** **** fully ******* *** ******* *********** ****** also ****** *** ********; ***** **** the ********* **** *********** ** *** distinguishing ******* ***** *** ********) *** a ********** ******** ******** ***** *** not **** ***** ******* * *** process ** ********* ** *** *.*. law *********** (****** *******/ ***********, ************* of ****** ****** ***** ****** ****** from ********* ****** - **** *** the *** **'* * *********** ***** puzzle).
** ** ** *** ***********, *********** only *** * *** ***** *********, and *** **** ******, ** ** is ******** **** **** ****** ** utilized.
******, *** **** ** ***** ***** in ***** ** ******* **********... ******* the ***** ******* ** ****** **** than * **** ** ******** ******** people **** ***** *****... ***?
***** *** ** ******* ****** *** meets * ******* *********, **** ****** and *** ************ ***** *** **** produce * ****.
”* ********** ******** ******** ***** *** not **** ***** ******* * *** process ** ********* ** *** ** law *********** (****** *******/***********, ************* ** arrest ***** ****** ****** **** ********* states”
** ******* *** **** ******** ** state, ******, *** ******* ********, *** updates ** *** **** ********* ** each ********* ******. ** * ****** is ****** ** ********, ** ********* won’t ** ** ********* ****** ****** and **** ** *** ********.
************ ****** *********** ** * **** up **** *********, *** *** ****** legal ********** **** ******** ********* ******** their ******** ** ********** ************* ***** only ** ******* *** *********** *************. An ************ ***** ** **** *** just ** *** ****** *** ** understand *** ******* *** ** *** in ***** ************* *********. **** *** intend ** ** ******* * ******* to * ********** ******** ** * municipal ****** (* ****** *** ********* as ** *******) *** *** ******** to **** ***** **** ******** ** the ************* ****** ********** *** ********* information **** ******* **** *** *******. BTW, **** **** * ******* ****** recognition ** ********** ****** ** ******* in *** *.*. ****** ****** *** political ********, **** ** ** *** NY.
**** ** ****** **** *** ******* of ***********'* ****** ********.