Business Trends and Outlook Survey, collection 19 Oct-1 Nov 2026: AI renamed "Super Intelligence", new headcount and hiring questions
A · Strong evidence Official US Census Bureau biweekly firm survey with published estimates and standard errors; this wave matters as a measurement break, and the new headcount and hiring items have no history yet.
Summary based on the full text·Original: about 15 minutes to read
In short
The US Census Bureau's Business Trends and Outlook Survey (BTOS) now words every AI question as "Super Intelligence or SI (formerly referred to as Artificial Intelligence or AI)". The change came in wave 202622 (collected 19 October-1 November 2026, published 5 November), the same wave that started a new sample cycle and added questions on staff use of generative AI and on AI's effect on headcount and hiring. Reported use barely moved, but "don't know" jumped, so the main US firm AI-adoption series breaks at this wave.
Key numbers
- 24.2%businesses using AI (now worded Super Intelligence) in the last two weeks, wave 202622, vs 24.6% prior wave
- 9.6% to 13.9%rise in 'Do not know' answers to the use question between the two waves
- 26.8%businesses whose staff used generative AI for work tasks (new item)
- 6.6% up / 2.5% downAI effect on total headcount in the last six months, among AI-using firms (87.3% no change)
- 2.5% up / 3.3% downAI effect on new hires in the last six months (90.2% no change)
As reported by the source.
English·Web page·opens census.gov
What it reports
- AI use in the last two weeks: Yes 24.2% (24.6%, 23.8%, 23.8% in the three prior waves); No 61.9% (65.8%); Do not know 13.9% (9.6%).
- Expected use in the next six months: 27.6% (28.6% prior wave).
- New: staff used generative SI for work tasks: Yes 26.8%, No 58.6%, Do not know 14.7%.
- New: effect of AI use on total headcount, last six months: increased 6.6%, decreased 2.5%, no change 87.3%, don't know 3.6%.
- New: effect on new hires: increased 2.5%, decreased 3.3%, no change 90.2%, don't know 4.1%.
- Wave 202622 is cycle 2, panel 1 of the sample, while the previous wave was cycle 1, panel 6.
How they got it
A voluntary biweekly online survey of US businesses; estimates are weighted shares of firms with published standard errors (for example 0.61% on the 24.2% use estimate, up from 0.27%). This digest reads the questionnaire PDF and the National.xlsx estimates file. The wording change is attributed on the BTOS about page to Executive Order 14434, as quoted by Simon Willison (2 November 2026); that JavaScript page itself could not be read.
Limits
- Wording, sample cycle and question set all changed in one wave, so the jump in "don't know" cannot be assigned to the wording alone.
- Headcount and hiring items are asked only of firms answering yes to AI or generative AI use, and are shares of firms, not of jobs.
- The saved questionnaire copy is labelled Content Year 5 with a 2027 field window; the live wording was confirmed in the data file.
- No bridge between old and new wording has been published.
Why it matters for companies
Anyone tracking AI adoption against US benchmarks, including Fed and academic work built on BTOS, should treat this wave as a new series. The new hiring item agrees with Australia's vacancy data and the Bank of England's survey: few firms report AI cutting hiring so far. Political renaming of a statistical concept is also a data-quality risk to plan for in any dashboard that relies on public series.
Summary based on the full text. Read the original for the full argument and data.Read the original on census.gov