Federal Data

The General Services Administration’s new rule, “Basic Safeguarding of Data within Large Language Model Artificial Intelligence Systems,” defines permitted and prohibited contractor actions involving government data in LLMs. Contractors should assess their AI systems and data practices as agencies seek greater control over government data and federal AI acquisition develops.

An analysis of 126,120 published comments found that 112,178—89%—opposed the Census Bureau’s proposed 2030 Census overhaul, including changes to demographic questions and which immigrants count toward congressional apportionment. Local-government and public-health stakeholders warned that revised population data could affect funding calculations and service planning, creating potential downstream implications for contractors that rely on census inputs.

Federal databases, surveys, and measurement systems are evolving in scope, access, and demographic coverage, complicating efforts to assess research outcomes over time. The issue points to potential demand for data governance, interoperability, analytics, and program evaluation support, but no specific solicitation, agency requirement, or funding amount is identified.