What happened
The U.S. Census Bureau published its “September 2026 HTOPS Household Pulse: September 9 – October 12” tables on November 12, 2026. The Bureau says the tables cover data collected through the Household Trends and Outlook Pulse Survey, or HTOPS, between September 9 and October 12, 2026. It describes HTOPS as continuing the Household Pulse Survey tradition of tracking national events with social and economic effects on U.S. households.
The release page organizes downloads into detailed tables, standard-error tables, and technical documentation. Its categories cover arts, employment, food sufficiency and security, health, housing, spending, Super Intelligence, and transportation. The linked detailed and standard-error files are Excel spreadsheets; the page identifies individual files as under 1 MB. The Bureau also says it reviewed the release for unauthorized disclosure of confidential information and approved the disclosure-avoidance practices, under reference CBDRB-FY27-0040.
The terminology change is explicit in several table titles. The Super Intelligence section includes tables for personal use of SI tools, frequency of personal use, attitudes toward use, concerns about use, and active avoidance or opting out. Each is described as covering the prior two months and as broken down by selected characteristics. Every relevant title adds the parenthetical clarification that the measure was formerly collected as artificial intelligence or AI.
The terminology also appears outside the dedicated section. An employment table covers use of SI tools for job-search activities during the prior two months. A health table covers trust in health or medical information from doctors and SI. The release provides corresponding standard-error spreadsheets for personal use, frequency, attitudes, concerns, opting out, job-search use, and health-information trust. The Bureau points readers to its source and accuracy statements for methodology and reliability information.
The page is therefore more than a single renamed chart. It applies the SI label across measures about household behavior, attitudes, work search, and health-information trust, while retaining a visible indication of the earlier AI wording. The published table structure also keeps uncertainty estimates adjacent to the reported measures, which matters for anyone extracting claims from subgroup comparisons rather than merely reading a label.
What the archive adds
New Runtime’s earlier coverage on November 2, 2026 noted that a Census Bureau adoption chart exposed a survey-question terminology shift and refreshed adoption trends through October 2026. This is a new fact on that known story: the September HTOPS release applies the SI wording across personal use, attitudes, concerns, opt-out, job search, and health-information trust tables.
Trend signal
This is a narrow but consequential extension of the terminology shift described in New Runtime’s earlier coverage. A public statistical instrument is now carrying “Super Intelligence” across more than one kind of household measure, rather than confining it to an isolated adoption chart. The release does not show that public behavior changed, or that the revised terminology improves measurement. This does not yet fit a pattern New Runtime tracks.
Why it matters
For teams using public surveys to estimate adoption, segment potential users, or support market claims, labels are part of the measurement instrument rather than cosmetic metadata. A switch from “AI” to “SI” can change what a respondent believes a question covers. A time series spanning that switch may therefore not be fully comparable even where the table topic and collection window appear familiar.
The practical response is to retain the question wording, field period, table identifier, and uncertainty estimate beside every figure cited internally or publicly. Treat a terminology update separately from movement in reported usage. When survey outputs inform product assumptions, market sizing, or trust claims, analysts should inspect the detailed and standard-error tables instead of relying on a category name or a chart-level trend.
Limits and uncertainty
The Census Bureau identifies the release date, field period, table subjects, downloads, disclosure review, and standard-error files. The published release page establishes that the label changed, but it does not independently demonstrate how respondents understood “Super Intelligence or SI” relative to “Artificial Intelligence or AI.” It also does not establish the size of any discontinuity that the wording change might introduce into comparisons with prior periods.
The estimates themselves require careful reading at the spreadsheet level, including their selected-characteristic breakdowns and standard errors. The Bureau directs readers to source and accuracy statements for methodological and reliability information. This Field Note does not independently evaluate the survey design, response behavior, or the statistical comparability of the renamed measures.
Reader takeaway
The September HTOPS tables are a new official release and a continuation of a measurement change, not standalone proof that household AI use rose or fell. If you use the figures, preserve the SI wording, survey period, table context, and standard errors—and do not assume that differently worded survey questions produce directly interchangeable trend lines.