> ## Content Index > Fetch the complete content index at: https://unhyd.com/llms.txt > Use this file to discover other available public pages before exploring further. # Maryland Business AI Benchmark: Broad Use, Modest Gains - URL: https://unhyd.com/article/maryland-business-ai-benchmark-adoption-workforce/ - Published: 2026-10-02T13:08:26.000Z - Updated: 2026-10-03T16:10:08.000Z - Description: A new state survey finds widespread AI use, limited integration and a workplace agenda centered on capacity rather than layoffs. - Author: Unhyd Editorial Staff - Tags: Business, AI, #sidebar-popular-posts, #unhyd-import, #home-featured-top Maryland’s new **Business AI Benchmark** offers a useful corrective to two easy assumptions about workplace AI: that adoption is still rare, or that it is already remaking companies from the inside out. The state’s survey finds broad use, but mostly at an early stage—and its respondents are more likely to expect employees to work with AI than to expect AI to eliminate their jobs. Governor Wes Moore’s office released the findings on October 1 at the Maryland Innovation Summit in Baltimore. The Maryland State Innovation Team said it surveyed nearly 300 senior decision-makers at organizations across the state between June and July. That makes the report a snapshot of what business leaders say they are doing and expecting, not proof that AI has caused a particular business or labor-market outcome. Still, the results give the state a more grounded starting point than broad predictions about automation. ## Maryland Business AI Benchmark: adoption is broad, integration is not According to the [governor’s office release](https://governor.maryland.gov/news/press-releases/governor-moore-convenes-maryland-innovation-summit-releases-new-maryland-business-ai-benchmark?ref=unhyd.com), 91% of respondents said their organization uses some form of AI. But 58% fell into the report’s basic-use category: standalone tools or AI features built into existing software, rather than deeper integration into business systems and workflows. That distinction matters. A team using an AI writing assistant or a search feature is not necessarily changing how work is assigned, reviewed or measured. Basic adoption can still be consequential, especially when a tool reaches many people. But it does not tell readers whether an organization has solved the harder questions: where human review sits, how output is checked, which data can be used, and who is accountable when a system is wrong. The productivity finding is similarly measured. Among regular AI users, 92% reported a positive effect on productivity. Most of those respondents, however, characterized the improvement as slight (70%), while 21% called it significant. That is a meaningful result without being a blank check for sweeping claims about return on investment. It suggests that ordinary efficiency gains are already visible to many users, while deeper operational change remains less common. ## Capacity, not cuts, is the immediate workforce signal The survey’s workforce results are more specific than the usual headline about jobs. Sixty-four percent of respondents said they planned to have existing employees do more with AI instead of hiring additional staff. Just 4% expected AI to reduce headcount. Two-thirds said they have a dedicated AI budget. The important qualifier is that doing more with the same workforce can change job quality even when it does not change the number of jobs. It can alter pace, performance expectations, training needs and the balance between junior and senior work. That is why survey-based expectations should not be read as a guarantee that layoffs will not occur. They do show that, for this group of Maryland decision-makers, the nearer-term plan is more often to expand capacity than to announce a direct reduction in roles. That focus is distinct from Unhyd’s recent coverage of [Data & Society’s Worker Lens research program](https://unhyd.com/article/data-society-ai-impact-on-workers-research/), which is designed to study how AI changes work inside particular industries. Maryland’s benchmark reports what senior business leaders say today; the research program will examine how workers experience deployment over time. The two approaches answer different questions, and both are needed. ## Why the state’s response is part of the story The benchmark was released alongside a broader state agenda that connects adoption to worker support and public safeguards. Maryland’s [AI policy principles](https://governor.maryland.gov/official-actions/ai-principles?ref=unhyd.com) call for worker and union input, training and transition support, while also setting out protections related to civil rights, privacy, safety and children. The immediate policy takeaway from the survey is practical. Businesses asked for capital and technical assistance first, followed by AI pilot programs and talent pipelines. In other words, the next constraint may be less about access to a model than about implementing it securely, training people to use it, and deciding where it belongs in a real workflow. Readers should watch for what follows the benchmark: whether the state publishes fuller methodology and recurring data, how small businesses are represented, and whether support programs measure outcomes beyond tool adoption. The most useful future reports will separate a one-off productivity boost from durable gains, and capacity growth from additional workload placed on employees. Maryland has established a baseline. The harder work is turning it into evidence about what changes for businesses and workers in practice. ## Sources - [governor.maryland.gov/news/press-releases/governor-moore-convenes-mar…](https://governor.maryland.gov/news/press-releases/governor-moore-convenes-maryland-innovation-summit-releases-new-maryland-business-ai-benchmark?ref=unhyd.com) - [governor.maryland.gov/official-actions/ai-principles](https://governor.maryland.gov/official-actions/ai-principles?ref=unhyd.com) - [thedailyrecord.com/2026/09/29/ai-wes-moore-summit-survey](https://thedailyrecord.com/2026/09/29/ai-wes-moore-summit-survey/?ref=unhyd.com)