MANCHESTER, NH – Just a small percentage of New Hampshire businesses use artificial intelligence, but the number that plan to in the future is on the rise, according to the Business Trends and Outlook Survey of the U.S. Census Bureau.
Granite State businesses come behind 40 other states and Washington, D.C., when it comes to using AI, according to the BTOS, which is an ongoing survey.
In New Hampshire, 3.5% of businesses in the survey used AI to produce goods or services in the previous two weeks, below the national 5.4% average. Some 4.9% of New Hampshire businesses expect to use AI in the next six months.
The survey also found that 99% of New Hampshire businesses that used AI in the previous six months did not experience an impact on employment.
The AI questions are part of the bigger BTOS survey, which is an experimental data product that tracks about 1.2 million U.S. businesses across all sectors, except farming, rotating to question about 200,000 of the businesses every two weeks. The AI segment was only available in the March 28 data release, the Census Bureau said.
The BTOS survey is the first to track business AI use in real time, rather than using older numbers, a report by the bureau’s Center for Economic Studies said.
“Tracking AI use by firms in a timely fashion has become even more pressing given the recent advances in Generative AI,” the CES report, which accompanied the release of the BTOS AI numbers, said. “Recent reports suggest ChatGPT has about 180.5 million users as of March 2024. However, it remains unclear how many businesses are currently using AI.”
The survey asked businesses if they used artificial intelligence in the two weeks before they were surveyed, as well as in the past six months, and if they plan to in the coming six months.
Businesses that said they used AI were asked what they used it for, and what impact it had on the business. Businesses that didn’t use it were asked if they planned to in the future, what applications they would use it for, or why they wouldn’t use it.
The CES study found that nationally, about 3.7% of businesses surveyed used AI when BTOS data collection began in September, and had risen to 5.4% by the end of February. It is expected to rise to 6.6% by fall.
The study found that AI use ranges from a low of 1.4% in the construction and agriculture sector to 18.1% in the information sector. Construction and agriculture expected use in the next six months is 1.5%, which information industry use is expected to rise to 21.5%.
Businesses that use AI often use it to substitute for worker tasks and equipment or software, nationally, according to the study, but that’s not something reflected in New Hampshire’s numbers. But like New Hampshire, few businesses nationally report reductions in employment because of AI use.
The study found that many firms undergo organizational changes to accommodate AI, particularly by training staff, developing new workflows, and buying cloud services/storage. Business that use AI also exhibit better overall performance and higher incidence of employment expansion, nationally, compared to other businesses.
AI-using businesses generally perform better than those that don’t use it, according to the study. They also more commonly experience employment expansions, rather than contractions when compared to businesses that don’t use it.
“The overall better performance of AI-using businesses is a fact that has repeatedly emerged from the analysis of other data, including data from the ABS [Annual Business Survey] and the BFS [U.S. Census Business Formation Statistics],” the study said.
The most common reason businesses don’t use AI, both in New Hampshire and nationally, is that it is not applicable to the business.
The study determined that about one in seven businesses that use AI now, may not in the future, but are using it on an experimental basis.
New Hampshire business AI use
New Hampshire’s 3.5% in AI use ranked it 42nd of all the states, Washington, D.C. and Puerto Rico in AI use in the two weeks before the survey; Vermont (3.1%) was 48th and Maine (2.3%) was 50th, followed by West Virginia (2.4%) and Mississippi (1.7%) was last. Colorado (7.4%), Washington D.C. (7.2%) and Florida (6.6%) were the top users.
While it’s not possible to cross-reference the publicly available BTOS overall numbers and the AI survey responses, in the general survey 13% of New Hampshire businesses reported excellent current performance, and 24.7% reported above-average current performance. Six months from now, 9.9% expect to report excellent performance, while 30.6% expect to report above-average performance.
New Hampshire’s responses to the supplemental AI questions in the BTOS survey:
In the last two weeks did you use AI in producing goods or services? 3.5% yes, 89.3% no, 7.2% did not know.
Why does this business not plan to use AI in the coming six months to produce goods or services? 85.8% AI is not applicable to this business, 5.7% AI is not a mature enough technology yet, 5% concerns about privacy/security, 4.6% lack of knowledge of the capabilities of AI, 3.7% other, 2.5% too expensive, 2.2% concerns about bias, 1.5% lack of skilled workforce.
In the last six months, what types or applications of AI did this business use in producing goods or services? 1.7% marketing automation, 1.4% speech/voice recognition, 1.3% virtual agents or chat bots, 1.2% large language models; 93% none.
Businesses that said they’d used AI, or planned to, were asked follow-up questions. Where several answers could be true, they were asked to pick all that apply.
In the last six months did this business use AI to perform tasks previously done by employees in producing goods or services? 19.5% yes, 72.1% no.
In the last six months, how many tasks previously done by employees were performed by AI? 82.5% a small number, no significant response for a moderate number or a large number.
In the past six months, did this business use AI to perform operations previously performed by existing equipment or software in producing goods or services? 79.8% no, no significant response for yes.
In the past six months, how did AI affect this business’s total employment? 99% it did not affect employment, no significant response for it did affect employment.
In the last six months, to use AI, what changes did this business make? 52.1 none, 22.7% developed new workflows, 18.2% trained staff to use AI.
During the next six months, do you think this business will be using AI in performing goods or services? 4.9% yes, 76.3% no, 18.8% do not know.
During the next six months, what types of AI applications will this business use in performing goods and services [for businesses that responded yes to the previous question]? 30.7% marketing automation, 24.5% text analytics, 24.3% virtual agents or chat boxes, 22.1% data analytics.
During the next six months do you think this business will use AI to perform tasks currently done by employees in producing goods or services? 40.1% yes, 56.5% no.
During the next six months, how many tasks currently done by employees will instead be performed by AI? 94.7% a small number, no significant response for moderate or large number.
During the next six months do you think this business will use AI to perform operations currently performed by equipment or software in producing goods or services? 30.3% yes, 50.3% no, 19.4% do not know.
During the next six months how do you think AI use will affect this businesses’ total employment? 94.4% will not change, no significant response for increase or decrease.
During the next six months, to use AI, what changes do you think this business will make? 54.9% train current staff to use AI, 41.9% develop new workflows, 25.1% none, 23.3% purchase cloud services or cloud storage, 23.2% change data management or data collection services.
Overall AI statistics
Some of the results of the BTOS supplemental AI survey nationally:
Common AI uses. The most common use of AI by businesses is marketing automation: 2.5% of all respondents, 28.4% of respondents that have used AI in the past six months, 36.5% of respondents that expect to use it in the next six months.
The second most common use is virtual agents or chat bots: 1.9% of all respondents, 21.6% of businesses that have used AI in the past six months, 28.2% of businesses that plan to use it in the next six months.
The most common use of AI reported by employees is data analytics: 6.6% employees who have worked at firms that reported using a specific application of AI in the past six months, 32.9% of those working for firms that did not respond “none” for use, 61.1% of those who expect to use AI.
The second most common type is robotics process automation: 5.9% that have used a specific application, 29.6% that did not respond “none,” and 31.5% that expect to use AI.
Employment. Employment changes by businesses in the past six months were: Overall increase 2.8%, AI users increase 6.5%; overall decrease 2.6%, AI user decrease 6.1%; no change overall 94.6%, no change by AI users 87.4%.
Business performance. A higher percentage of businesses that use, or expect to use AI, reported excellent or above-average performance than those that don’t, or don’t expect to. The CES study emphasizes that the survey shows no causal link between AI use and performance. Still, 38.9% of AI users reported “excellent” or “above average” performance in the weeks before the BTOS survey, 31.4% of nonusers report excellent or above average. Some 48.1% of AI users expect excellent or above average future performance; 30.9% of nonusers have the same expectations.
Size of businesses using AI. AI, nationally, is more frequently used in businesses with very large numbers of employees, 250 or more, or very small, 1-4 employees. Of firms with 250 or more employees, 7.2% use AI, with 9% of all workers in firms that use AI. About 11% of businesses that size expect to use it in the next six months. About 5.5% of businesses with 1-4 employees use AI, with 7% expecting to use it in the next six months.
Age of businesses using AI. Businesses that are less than five years old are most likely to use AI, at 6.5% and most likely to expect to use it in the next six months, at 9%. The older a business, the less likely it is to use AI, with 3.9% of businesses more than 21 years old nationally using it, and 4.2% expecting to use it. That said, of businesses that use AI, the oldest businesses are almost as likely to use it as the youngest, and 10% of both say they expect to in the next six months.
‘A rich snapshot’
The CES study cautions that there are challenges measuring AI business use that should be taken into account when looking at the findings.
“Firms may use AI in incidentally or insignificant ways, and not report such use, as AI may not be a part of their core processes or products,” the CES study said. “Others may rely on AI embedded in systems or services provided by third parties; in which case they are unknowing users of AI. The estimates produced from business responses to the BTOS may not include these incidental and/or embedded uses of AI. Still, we think that the BTOS statistics are highly relevant as they arguably reflect more direct and significant use of AI for business purposes.”
Even so, the study notes that the BTOS survey results provide “a rich snapshot of the patterns of recent and future AI use by firms.”
It shows that the use of AI in producing goods and services remains low, it is growing, and is associated with better business performance overall, but not leading to declines in employment in the present period and near future, it says. “Whether these patterns will change as the diffusion of AI progresses remains to be seen,” it adds.
The findings are consistent with the view that, although AI has advanced rapidly in recent years, its use by businesses is still in its early days.
“Many businesses do not yet see AI as being relevant for their business activity or do not yet have the knowledge to implement AI in their business activity,” the study says. “Training and complementary investments in changing business organization in various ways are common activities of users of current and future users AI, so non-users face these challenges in adopting AI.”