0
Skip to content

How Businesses Are Actually Using Gen AI: Boosting Capacity Over Cutting Jobs

Woman and robot working on a computer. AI notetaker in progress.New Federal Reserve data shows companies are taking a cautious, step-by-step approach to generative AI — focusing on employee productivity rather than mass layoffs, even as fresh IP and regulatory risks emerge.

07/23/2026 4:00 P.M.

While generative AI adoption is expanding at a significant pace, new survey data from the Federal Reserve Bank of St. Louis reveals that most regional businesses are still in the early, experimental stages of implementation — focusing on expanding capacity rather than cutting headcount.

Most firms are treading carefully rather than overhauling operations overnight:

  • 34% of surveyed businesses report regular AI tool use among a small subset of employees.
  • 25% are currently testing or running pilot programs.
  • 11% report no AI adoption whatsoever.

Rather than sweeping systemwide rollouts, even major organizations (like regional health care networks) report integrating AI into workflows step-by-step.

For firms using AI, the primary dividend isn’t headcount reduction — it’s capacity expansion. Early adopters in professional services and knowledge sectors report significant productivity gains by automating routine and administrative tasks:

For the roughly one-third of firms holding off or limiting adoption, the primary roadblocks include:

  • Skill & Infrastructure Gaps (38%): Lack of internal technical expertise, required data architecture, or employee training.
  • Perceived Utility (34%): Doubts about whether available AI tools currently address core business needs.
  • Risk & Compliance Concerns: Financial institutions specifically cited risk management hurdles — including system controls and AI “hallucinations” — as major barriers to broader deployment.

What It Means for the Workforce

Despite widespread anxiety about AI-driven layoffs, short-term labor dynamics appear stable:

  • 49% of firms expect zero effect on overall staffing over the next 12 months.
  • 18% anticipate shifting skill requirements for existing roles rather than net job cuts.
  • About 20% foresee a slight reduction in staffing needs (mainly in transactional roles, such as automated order-taking).

The study shows that AI is acting as a force multiplier rather than a replacement strategy. In the near term, the biggest competitive threat isn’t that AI will take jobs — it’s that AI-enabled firms will outpace non-adopting competitors by doing significantly more work with the same team.

Brownstein Hyatt Farber Schreck recently shared in a client alert, “When AI Models Learn from Employees: The Next Trade Secret Battleground,” that while generative AI dramatically boosts efficiency and grants competitive advantages by streamlining everyday tasks like summarizing documents, coding, taking meeting notes, and analyzing institutional data, heavy organizational reliance on these tools creates a subtle, growing threat to trade secret security that many companies have not yet addressed.

“Historically, companies have focused their protective efforts on preventing the physical or digital removal of confidential materials,” the alert states. “Trade secret disputes have typically revolved around identifiable acts such as downloading files, copying source code repositories, forwarding sensitive emails or taking customer lists. These were discrete, traceable events tied to tangible information. AI systems, however, complicate this framework by enabling a more diffuse and less visible form of knowledge transfer.”

Brownstein notes that AI allows companies to capture and scale individual expertise into reusable outputs and workflows, but this makes proprietary knowledge easier to share, adapt, and move outside traditional controls.

When employees leave, proving trade secret theft becomes far harder. Disputes will increasingly center on whether AI was used to recreate a former employer’s confidential methods, complicated by the fact that AI outputs lack clear provenance and are difficult to trace, according to the Brownstein alert.

ACA’s Take

In the accounts receivable management industry, collection firms are starting to integrate generative AI voice and text agents to automate outbound consumer outreach. While these tools offer efficiency and operational advantages, media reports have documented instances where automated systems contacted consumers to collect on bills that had already been paid or disputed, ACA International previously reported.

Over the past year, federal and state regulators have also signaled heightened scrutiny regarding automation.

State-level framework developments show that lawmakers are actively building guardrails around automated decision-making and consumer communication. Additionally, federal watchdogs continue to monitor how technology intersects with consumer protection statutes.

Do you have questions about how your peers are using AI compliantly? Ask them in ACA’s members-only online community, The Exchange. Recent conversations have discussed auditing AI agents, managing AI risk, and more.

Related Content from ACA:

Remember, subscribe to ACA Daily and Member Alerts under your My ACA Assistant profile when logged in to acainternational.org.

If you have executive leadership updates or other member news to share with ACA, contact our communications department at comm@acainternational.org. View our publications page for more information and our news submission guidelines here.

One moment please...

Share Profile

This site utilizes cookies as included in our Privacy Policy and User Agreement.