surveillance_spending

A DISCOVERY INDEX of surveillance and police-technology dollar figures found in local government meeting records across 130+ named vendors — NOT a table of confirmed spending, and NOT safe to SUM. An audit on 2026-07-29 found aggregate views were adding amounts that must not be added: the same procurement recurs at committee, council, approval, amendment and renewal stages; "not to exceed" figures are authorization ceilings, not payments. Use the canned query "the CONSERVATIVE total" for any number you intend to publish, and "data-quality report" to see what is flagged. Columns added for this: `amount_type` (stated / ceiling / proposed / recommended / estimated), `dup_rank` (>1 = redundant copy), `geo_suspect` (city disagrees with the source link), `anomaly` (implausible figure), `state_norm` (normalized state). Vendors covered include ALPR (Flock, Vigilant, Genetec), body cameras/tasers (Axon, WatchGuard, Utility, Getac), gunshot detection (ShotSpotter/SoundThinking), face recognition (Clearview AI, DataWorks Plus), fingerprint biometrics (Idemia AFIS/Livescan, filed separately under `sv:biometrics`), data fusion and OSINT (Fusus, Peregrine, LexisNexis Accurint, Thomson Reuters CLEAR), phone forensics (Cellebrite, GrayKey, Magnet), drones (Skydio, BRINC, DJI), cameras/VMS (Verkada, Avigilon), cell-site simulators, CAD/RMS records systems, photo enforcement, jail-call and student monitoring. Every row carries a VERBATIM quote from the minutes and a link to the source document, and a row only publishes if that quote was found character-for-character in the source text. READ THE `ledger` COLUMN: `sv:<category>:<vendor>`. `radio_infra` (two-way radios), `rms` (records/CAD software), `biometrics` (fingerprints) and `not_surveillance` are NOT surveillance sensors — exclude them before quoting any surveillance figure.

Data license: CC BY 4.0 · Data source: MyTown

1 row where ledger = "sv:fusion:thomson_clear", state = "MN" and vendor = "Thomson Reuters CLEAR" sorted by amount descending

✎ View and edit SQL

This data as json, CSV (advanced)

Suggested facets: spend_date (date)

vendor 1

  • Thomson Reuters CLEAR · 1 ✖

state 1

ledger 1

  • sv:fusion:thomson_clear · 1 ✖
Link rowid vendor city state spend_date product amount ▲ quote source_document mytown_record ledger amount_type dup_rank geo_suspect anomaly state_norm
3683 3683 Thomson Reuters CLEAR Duluth MN 2021-09-13 CLEAR fraud software (approve contract) 3312.0 RESOLUTION APPROVING AN AGREEMENT BETWEEN THE CITY OF DULUTH AND THOMSON REUTERS FOR TWO ONLINE SUBSCRIPTIONS TO CLEAR FOR GOVERNMENT FRAUD SOFTWARE FOR AN INITIAL TERM OF THREE YEARS AT AN ANNUAL COST OF $3,312.00 WITH ANNUAL INCREASES IN YEARS TWO AND THREE OF FIVE PERCENT. https://duluth-mn.legistar1.com/duluth-mn/meetings/2021/9/1943_M_City_Council_21-09-13_Minutes.pdf https://mytown.theboringparts.com/city/duluth-mn/2021/#m278914 sv:fusion:thomson_clear stated 1 0 0 MN

Advanced export

JSON shape: default, array, newline-delimited

CSV options: 

CREATE TABLE surveillance_spending (
  vendor TEXT, city TEXT, state TEXT, spend_date TEXT, product TEXT,
  amount REAL, quote TEXT, source_document TEXT, mytown_record TEXT, ledger TEXT, amount_type TEXT, dup_rank INTEGER, geo_suspect INTEGER, anomaly INTEGER, state_norm TEXT);
CREATE INDEX ix_surv_vendor ON surveillance_spending(vendor);
CREATE INDEX ix_surv_state ON surveillance_spending(state);
CREATE INDEX ix_surv_amount ON surveillance_spending(amount);
CREATE INDEX ix_surv_ledger ON surveillance_spending(ledger);