Canonical segment URLs exist only for observed category/shape pairs.
Machine Relations Index
Observed shapes1 Published1 Collecting0 ReleaseNovember 2, 2026
Legacy News Topics · Machine Relations Index category
Which sources AI engines cite for Legacy News Topics questions
Vendor-owned sources earn 37.0% of classified citations in Legacy News Topics answers (3,105 of 8,382), as of November 2, 2026.
In the November 2026 release of the Machine Relations Index, these are the sources AI answer engines cited most often for the one Legacy News Topics question shape it observes, each ranked on that shape's own observed runs. This category has 1 observed question-shape segments in the current public release. Each segment keeps its own observed-run denominator and dates; this page does not average rates across shapes. Observed runs are answer runs, not questions: every segment repeats a small set of distinct questions across engines and dates, so a run count is not a count of independent questions.
Top sources by question shape
The sources engines cite most, shape by shape
A rate is the share of the shape's observed answer runs that cited the source; rank is the source's place on that shape's full leaderboard.News-driven citations
Example pattern: recent legacy news topics news. 2,066 observed runs across 54 dates.
| Rank | Source | Source role | Citation rate | Cited runs |
|---|---|---|---|---|
| #1 of 4,017 | Community and social platform | 10.21% | 211 of 2,066 | |
| #2 of 4,017 | Medium | Community and social platform | 9.20% | 190 of 2,066 |
| #3 of 4,017 | Gartner.com | Analyst and consulting research | 8.81% | 182 of 2,066 |
| #4 of 4,017 | G2 | Market and company database | 8.03% | 166 of 2,066 |
| #5 of 4,017 | Community and social platform | 7.79% | 161 of 2,066 |
Published segments met the release evidence floor.
Collecting segments expose sample progress, not mature leaderboards.
Methodology v2.1.
Question shapes
Separate denominators by segment
| Question shape | State | Observed runs | Distinct questions | Run dates | Distinct cited domains | Canonical segment |
|---|---|---|---|---|---|---|
| News-driven citations | Published | 2,066 | 85 | 54 | 4,017 | Open leaderboard |
By answer engine
Whole-window rates on each engine
Each link ranks sources with a published Legacy News Topics segment by that source's citation rate on one engine across the whole window. It is not a rate inside Legacy News Topics alone. A collecting engine has no published ranking for this category.Cite this
Citation
Machine Relations Index (MRI), Machine Relations Research — machinerelations.ai. Legacy News Topics, observed Machine Relations Index category, 2026-05-10 to 2026-11-02. Release mri_score_v2.1+2026-11-02+f9aaeca0db15. https://machinerelations.ai/index/categories/legacy-unmapped. CC BY 4.0.APA
Parrott, J. (2026). Legacy News Topics in the Machine Relations Index (Version mri_score_v2.1+2026-11-02+f9aaeca0db15) [Data set]. Machine Relations Research. https://machinerelations.ai/index/categories/legacy-unmapped https://doi.org/10.5281/zenodo.23652423BibTeX
@misc{parrott2026mrilegacyunmapped,
author = {Parrott, Jaxon},
title = {Legacy News Topics in the Machine Relations Index},
year = {2026},
version = {mri\_score\_v2.1+2026-11-02+f9aaeca0db15},
publisher = {Machine Relations Research},
url = {https://machinerelations.ai/index/categories/legacy-unmapped},
doi = {10.5281/zenodo.23652423},
license = {CC-BY-4.0},
note = {Machine Relations Index (MRI), Machine Relations Research — machinerelations.ai}
}Release and boundary
No pooled category score
Published November 2, 2026. The observation window runs May 10, 2026 to November 2, 2026. Methodology v2.1.
Use the data
curl -fsSL "https://machinerelations.ai/data/machine-relations-index.json" -o machine-relations-index.json
curl -fsSL "https://machinerelations.ai/data/mri-release-manifest.json" -o mri-release-manifest.json
curl -fsSL "https://huggingface.co/datasets/machine-relations/machine-relations-index/resolve/main/domains.csv" -o domains.csv
curl -fsSL -L "https://doi.org/10.5281/zenodo.23652423" -o mri-zenodo.htmlimport json
from urllib.request import urlopen
index_url = "https://machinerelations.ai/data/machine-relations-index.json"
with urlopen(index_url) as response:
index = json.load(response)
print(index["generated"], index["total_domains"])This page is a directory of observed segments. A domain's overall citation rate is not a category rate, and MRI does not publish an averaged category score in this slice.