Using an LLM pipeline, we assessed how well each document in CSET Emerging Technology Observatory's AGORA dataset covered industrial sectors based on the North American Industry Classification System. Each document may cover multiple sectors. See the pilot blog post for discussion of the methodology.
Insights from analysis of Sectors Governed:
- Concentration in public and research-oriented sectors. The sectors most frequently referenced in the AGORA dataset are concentrated in the public sector and research-oriented domains, including Public Administration (excluding National Security), Scientific Research and Development, Information, and National Security.
- Human related sectors receive less coverage. Sectors more directly tied to everyday life and economic activity receive substantially less attention. The least frequently covered sectors include Accommodation, Food, and Other Services, Management, Administration, and Support Services, and Real Estate and Rental and Leasing.
- Policy implication. This distribution suggests that AI governance documents within the AGORA dataset have prioritized institutional, technical, and security-oriented contexts, while governance of consumer-facing and labor-intensive sectors—where societal and economic impacts are more immediate—has received comparatively less attention.
Interactive Chart
Explore the chart using the dropdown filters, and by clicking on category names (Jurisdiction, Authority, Legislative Status etc) to see distributions within each category or click on the preset example configurations below the chart.
[ { "label": "Sectors with good coverage, broken down by Jurisdiction", "filters": { "coverage": ["3"], "stackBy": "jurisdiction" } }, { "label": "Sectors with good coverage, broken down by Authority", "filters": { "coverage": ["3"], "stackBy": "authority" } },
{ "label": "Sectors with good coverage, broken down by Legislative Status", "filters": { "coverage": ["3"], "stackBy": "legislativeStatus" } }, { "label": "Sectors with good coverage in enacted documents only, broken down by Jurisdiction", "filters": { "coverage": ["3"], "activity": ["Enacted"], "stackBy": "jurisdiction" } }, { "label": "Sectors with good coverage in documents targeting Private sector", "filters": { "coverage": ["3"], "appliesTo": ["Private sector"] } }
]
Important context for interpreting these results:
The LLM classifications are based on the approximately 1000 documents in the AGORA dataset, which is predominantly composed of U.S.-origin English language government proposed documents, the majority of which are federal-level.
Coverage patterns described therefore reflect the priorities and framing conventions of this particular corpus and should not be taken as representative of the global AI governance landscape. We also found LLM classifications to exhibit some biases including over-attribution of coverage when governance-related language is present. Coverage scores should be taken as indicative of broad patterns rather than precise measurements.
%20(250%20x%20100%20px).png)
.png){kind=small}
