US school district web census

Census fetched 2026-09-30 · population: NCES Common Core of Data, Local Education Agency Universe Survey directory 2024-25 (agencies with a website address) · datasets (CC BY 4.0)

The US school district web census, September 2026: what 15,896 local education agencies publish for families, security researchers and AI agents

Every local education agency in the NCES Common Core of Data directory with a website address (15,896 school districts and other agencies) was fetched once over HTTPS on 2026-09-30, with robots.txt, security.txt (both locations), llms.txt and the A2A agent card. 14,653 answered with a website.

Reachable over HTTPS

92.2%

14,653 of 15,896 school districts

security.txt

27.3%

4,003 of reachable sites; 3,889 complete and current

Accessibility statement linked

21.1%

3,092 homepages

Privacy policy linked

11.4%

1,677 homepages

Block GPTBot

555

538 block ChatGPT-User; 546 ClaudeBot

llms.txt

566

0 publish an A2A agent card

What we found

By agency type

Agency typeIn listReachablesecurity.txtvalidprivacy linkaccessibility linkHSTSblocks GPTBotblocks ChatGPT-Userllms.txtagent card
Regular public school district that is not a component of a11,77310,942 (92.9%)3,715 (34.0%)3,6129.3%22.7%80.0%3713611880
Independent charter district3,1012,807 (90.5%)137 (4.9%)13117.9%13.8%50.7%1541513220
Service agency446417 (93.5%)88 (21.1%)8415.8%20.6%68.1%1614270
Specialized public school district254240 (94.5%)38 (15.8%)3712.5%19.2%63.8%87150
Other local education agency11892 (78.0%)10 (10.9%)1025.0%33.7%60.9%3280
State operated agency9583 (87.4%)4 (4.8%)430.1%43.4%56.6%3320
Supervisory union5434 (63.0%)4 (11.8%)423.5%32.4%50.0%0030
Regular public school district that is a component of a supe5336 (67.9%)7 (19.4%)725.0%30.6%69.4%0010
Federal operated agency22 (100.0%)0 (0.0%)00.0%0.0%50.0%0000

By state

States with at least 20 school districts in the list; percentages are of reachable sites.

StateIn listReachablesecurity.txtprivacy linkaccessibility linkHSTSblocks GPTBot
CA1,6571,523 (91.9%)14110.5%19.0%60.4%72
TX1,2251,174 (95.8%)3297.0%30.5%80.0%21
OH953861 (90.3%)18712.7%14.5%72.7%106
NY944863 (91.4%)16120.2%32.3%80.8%47
IL844765 (90.6%)34511.5%14.2%80.3%8
MI768720 (93.8%)22811.3%28.5%63.2%14
PA723675 (93.4%)1138.6%14.1%72.1%37
NJ673650 (96.6%)1634.9%14.8%77.8%54
AZ569502 (88.2%)6819.1%23.7%63.7%32
MN569545 (95.8%)10810.5%8.1%74.1%12
MO547522 (95.4%)2159.2%21.1%82.4%13
OK543517 (95.2%)2267.9%39.7%89.9%4
WI466340 (73.0%)996.8%15.9%55.0%2
MA378358 (94.7%)6112.8%24.3%65.4%15
IN375338 (90.1%)889.8%22.5%68.9%9
IA334319 (95.5%)10720.7%26.0%64.6%5
KS314304 (96.8%)13715.5%28.3%86.2%2
AR303289 (95.4%)1748.7%20.1%90.0%0
MT260219 (84.2%)838.7%19.2%63.9%19
NC252214 (84.9%)4411.2%12.6%73.4%9
WA228219 (96.1%)647.8%23.7%81.7%2
OR208196 (94.2%)5111.2%32.1%82.1%2
CO205198 (96.6%)6522.2%24.7%66.2%10
CT191182 (95.3%)2316.5%23.6%73.1%8
ND191176 (92.1%)998.0%6.3%76.7%7
NM184171 (92.9%)5012.3%22.8%77.8%3
ME177170 (96.0%)7213.5%7.1%70.6%1
SD177164 (92.7%)956.1%6.7%77.4%5
KY176157 (89.2%)653.8%15.9%93.0%3
VA161154 (95.7%)5411.7%16.9%75.3%5
AL153143 (93.5%)2611.9%32.9%90.9%3
MS149129 (86.6%)218.5%15.5%85.3%1
UT144134 (93.1%)1532.1%16.4%40.3%1
TN137119 (86.9%)288.4%23.5%76.5%1
FL8568 (80.0%)2214.7%35.3%80.9%1
SC8277 (93.9%)337.8%7.8%89.6%0
GA8067 (83.8%)224.5%10.4%73.1%5
VT7150 (70.4%)1022.0%32.0%66.0%0
DC7070 (100.0%)218.6%7.1%44.3%1
NH6762 (92.5%)1012.9%4.8%61.3%1
WY6159 (96.7%)2210.2%27.1%67.8%4
AK5453 (98.1%)2020.8%9.4%75.5%0
WV4944 (89.8%)392.3%0.0%90.9%0
DE4140 (97.6%)67.5%20.0%52.5%7
MD2524 (96.0%)916.7%12.5%79.2%1
NV2017 (85.0%)323.5%17.6%82.4%0

Named lists

Largest school districts without a security.txt (first 60 of 300)

Sites with an expired security.txt (5)

Sites publishing an llms.txt (first 60 of 566)

Sites telling ChatGPT-User and Claude-User not to open pages for users (first 60 of 284)

Sites blocking GPTBot (OpenAI's training crawler) (first 60 of 300)

Homepages that redirect to another domain (first 60 of 300)

Sites with a Content-Signal line in robots.txt (44)

Largest school districts without an accessibility link on the homepage (first 60 of 300)

What school districts are expected to publish

Look up any school district

Questions

Why do so many school districts publish a security.txt?

4,003 of the 14,653 reachable district sites (27.3%) serve a security.txt, but a spot check of ten of them found every file supplied by the district's website vendor (Apptegy, Contact: security@apptegy.com), not written by the district. Read the figure as vendor coverage of vulnerability disclosure, not district practice.

How many school district websites publish a security.txt?

4,003 of the 14,653 reachable district websites (27.3%) publish a security.txt; 3,889 are complete and current.

Do school district websites link an accessibility statement?

21.1% of reachable district homepages link an accessibility statement and 11.4% link a privacy policy. The ADA Title II web rule reaches districts serving 50,000 or more people on April 24, 2026 and smaller districts on April 26, 2027.

How many districts block AI crawlers?

555 district websites block GPTBot, 546 block ClaudeBot and 538 block ChatGPT-User in robots.txt; 279 disallow every bot; 566 publish an llms.txt.

Which districts are included?

Local education agencies in the NCES Common Core of Data directory that list a website address: 15,896 agencies on 2026-09-30. Agencies without a website in the directory are not included.

Can I reuse the data?

Yes. The findings, the CSV and every per-district JSON record are published under CC BY 4.0; cite Agent Exchange and link the report.

Method and limits

Population: NCES Common Core of Data, Local Education Agency Universe Survey directory 2024-25 (agencies with a website address) with each agency's type, city, state and number of operational schools as published for 2024-25. Many district sites run on a handful of website vendors; vendor-supplied files such as security.txt are counted as published, and the FAQ says so.. Fetch: one HTTPS request to the apex (then www.), following redirects, 8-second timeout, user agent agentexchange-census/1.0; then robots.txt, /.well-known/security.txt, /security.txt, /llms.txt and /.well-known/agent-card.json on the final host. "Reachable" means any HTTP response over HTTPS, including 403s from bot protection. Homepage links are detected from anchor targets and visible text, so a policy that exists but is not linked from the homepage is not counted. A robots.txt "blocked" verdict means the agent's own group (or the * group) disallows the whole site; "partly blocked" means some paths are disallowed. One snapshot per domain; slow sites and sites that block automated requests are counted as unreachable or reachable-with-error as the server answered. Per-domain records: lookup; downloads: CSV, NDJSON and JSON (CC BY 4.0).

Run one of these sites?

$49 full report or $99 monthly monitoring · Hosted accessibility feedback and privacy request desk

Other censuses

Government web census (.gov) · US bank web census · US college web census · US public library web census · Top 100,000 websites census · AI crawler census (top 10,000 sites)