Incorporating AI within business strategy and key processes has quickly become essential.
Key takeaways
Companies are shifting how they use AI, focusing on embedding it in everyday operations.
AI success depends on many factors, including defined objectives, governance and skill depth.
Artificial intelligence is at the forefront of a rapidly evolving middle market business landscape, transforming processes, driving innovation and creating long-term value. However, successfully capitalizing on AI requires strategic planning, clearly defined business use cases, organizational AI readiness and building trust in AI-driven outcomes through proof of concept (POC) testing and phased implementation. In addition, decision makers must establish a practical roadmap to scale AI responsibly and effectively.
In 2026, companies are shifting from mainly using AI as stand-alone tools to embedding AI as a core element of everyday operations. With that transition, many strategies that were focused on chat-based tools are evolving to include autonomous agents with seamless integration into critical workflows.
Several key trends are shaping the use of AI solutions in 2026, including:
The rise of agentic AI
Companies are moving from passive chatbots to AI agents that can execute complex tasks and perform autonomous decision making with limited human intervention. These agents can increase productivity, enhance the customer experience with consistent, frictionless interactions, and improve security operations by independently identifying and addressing threats.
Physical and multichannel solutions
AI interactions are becoming more common in physical environments and across multiple data types. AI tools can now analyze and interact with text, audio and video in real time, helping companies enhance user experiences and operations.
Hyper-personalization for deeper impact
AI tools have become more personalized to address the needs of specific industry sectors and research demands.
Technical and operational evolution
Large language models (LLMs) are still prevalent, but small language models (SLMs) that run locally—on mobile devices or edge systems rather than in the cloud—are becoming popular for reasoning tasks.
Ethical and trust concerns
As AI agent use increases, companies are focusing on using AI responsibly, building trust in output and enhancing complex reasoning in settings where precision is necessary.
Demonstrating how quickly AI use has advanced in the middle market, the RSM Middle Market AI Survey 2025: U.S. and Canada found that 91% of middle market executives said their organizations are either formally or informally using AI in business practices. But 53% whose organizations have implemented generative AI believe they were only somewhat prepared to do so and 70% reported needing outside help to get the most out of their AI solutions.
Furthermore, 92% of executives experienced challenges with implementation and 62% said generative AI has been harder to implement than expected. Given the rapid growth of AI and the implementation challenges that arise, companies must establish a well-defined, result-oriented AI strategy and roadmap.
In RSM US LLP’s recent webinar Inside the middle market: 2025 AI trends, challenges and executive insights, Principal Paul Seckar and Director Robbie Beyer, along with Danny Buie, Vice President of IT, Tyler Technologies and Dan Dwight, President and CEO, Cooley Group, provided data-backed insights and executive perspectives on moving from experimenting with AI use cases to creating meaningful impact.
Below, we look at their key considerations for developing an AI adoption strategy, outlining the challenges, opportunities and methodologies needed to drive growth, scale responsibly and gain a competitive advantage.
Importance of a defined AI strategy and data roadmap
To be truly effective, an AI strategy must align with a company’s overall strategy and long-term goals. Measures to support alignment include:
- Establish data governance: High-quality data is critical to successfully deploying AI solutions. Start with a few narrow, high-quality use cases to build trust with stakeholders across the company.
- Build an understanding of AI: If you are starting from square one, learn how AI works and how it applies to your business pain points. Setting expectations early helps identify meaningful use cases and the data needed to support them.
- Define clear objectives: Identify measurable goals to understand success factors and increase efficiency. Operational performance improves in direct proportion to the strength of strategic planning and data management.
AI adoption should be treated as a core business strategy, centered around how an organization is transforming and how it wants to operate effectively with data as an enterprise asset. It certainly helps gain valuable insights for informed decision making.
Paul Seckar, Principal, RSM US LLP
AI implementation challenges
Common AI implementation challenges include data quality, data privacy and security, and insufficient internal resources. In the 2025 survey, among respondents who experienced AI implementation issues, 41% expressed concerns about data quality, the top problem companies faced. In addition, 39% of respondents who said their organization was unprepared for their AI implementation cited a lack of in-house expertise as their top issue.
Companies can leverage several strategies to address implementation challenges, including:
- Conduct multiple employee training sessions on AI basics, data privacy and data security to prevent the misuse of public AI tools, reinforcing the importance of high-quality data.
- Develop an AI use case for a specific task, such as forecasting or customer churn analysis.
- Create enterprise-wide use cases to support investment in long-term data governance, efficiency, quality and scalability across all functions and processes.
“For some businesses, it makes sense to focus on a single high-value use case such as forecasting rather than attempting an enterprise-wide data overhaul,” says Beyer. “The right approach depends on business pain points, current challenges and project appetite. AI should never be implemented just because it is a trendy technology.”
The availability of skilled internal AI talent is an ongoing challenge in deploying and maximizing the value of AI investments. Therefore, many companies are turning to trusted third parties to supplement their internal teams to develop and implement successful AI strategies .
Buie shared his experience developing a comprehensive AI talent strategy at Tyler Technologies. “Internally, we couldn't get where we needed to be initially because we didn't have the skills—we were going to have to grow them,” he says. “So we adopted a model with RSM that I call ‘service as a service,’ where we could utilize RSM’s advanced machine learning talent and some other AI skill sets. While internally we're growing and building and upskilling our own team members, we weren't going to get there fast enough. This is where we augmented our team with RSM's knowledge—as our team builds up that skill.”
Internally, we couldn't get where we needed to be initially because we didn't have the skills—we were going to have to grow them. So we adopted a model with RSM that I call ‘service as a service,’ where we could utilize RSM’s advanced machine learning talent and some other AI skill sets.
Danny Buie, Vice President of IT, Tyler Technologies
How to increase AI adoption
To enhance overall efficiency and garner long-term value from AI deployment, organizational readiness is essential. A key readiness component is AI and data literacy, which can be developed through training sessions and office hours and include identification of usage opportunities and associated risks. Socializing elements of the data and AI strategy with the enterprise and linking it to core elements of the existing business strategy is another means of boosting AI and data literacy.
“Successful AI implementation must begin with a clearly defined business problem, not with the technology itself,” says Beyer. “Teams should work backward from the pain point to understand data requirements, evaluate data quality, assess available tools and determine how users will adopt and scale the solution.”
Buie echoed the importance of prioritizing AI use cases. “For the use case that we went after, we did seek additional executive support because we’re a publicly traded company and we were concerned about our public auditors using AI on our data,” he says. “We were able to get sponsorship for this specific use case to use machine learning to get ahead of the auditors with our financial data and financial transactions. We were able to get funding and then used data that we know is reviewed and audited and reconciled. All of those things helped build trust in the process and the solution.”
While early returns may be hard to quantify, organizations that stay committed may see significant long-term value. To prioritize AI initiatives, leaders must evaluate their current business challenges and address them accordingly. Ultimately, starting with business pain points and applying a structured value lens help identify compelling AI initiatives that are easy to understand and scale.
AI implementation risks and rewards
AI is rapidly transforming middle market business processes, but companies must follow a thoughtful process to implement the right solutions and get their expected value.
Companies should follow a “stage-gate” process that begins with a problem statement, followed by data collection and analysis. Leadership should then evaluate all solution options, including AI tools, based on their risks and benefits. This disciplined approach ensures that AI is used intentionally and only when it adds real value.
In addition, to maximize the returns on digital investments, companies should treat AI like any other technological tool and not assume it solves all problems.
“Many organizations treat AI like a hammer looking for nails, which can lead to poor results and unnecessary costs,” says Beyer. “Instead, teams should determine whether AI is truly the right tool for a specific business problem. In many cases, especially in finance functions like period close, traditional tools or simpler automation may be more effective. AI can supplement workflows, but it is not a silver bullet.”
Starting with a POC is a key step in building success and mitigating risk. A proven strategy is to adopt AI through a phased approach—launching a pilot within a single business process or geographic location and learning from the experience before scaling it across other processes or locations.
In addition, POCs help companies see AI in action on their own data. While education and theory can help predict what “should” happen, real dashboards and real outputs enable decision makers to clearly understand how AI fits into their workflows at a granular level. POCs also allow teams to learn quickly, validate assumptions and decide whether to scale a solution or pivot to something more valuable.
A POC allows leaders to start seeing and planning for critical things early on as opposed to jumping all the way into the deep end trying to conduct enterprise-wide rollout. It also supports cost and budget optimization. As you scale from a POC and experimentation into production, you optimize opportunities to bring repeatability into the process, making it easier to roll out and scale across the organization.
Robbie Beyer, Director, RSM US LLP
Key AI priorities and possibilities
Organizations navigating AI adoption have a growing number of options and strategies to consider, including capitalizing on technology ecosystems with robust AI capabilities that continue to develop and advance over time.
For example, many Microsoft solutions like Fabric and Power BI are already embedded with advanced AI functionality. Fabric is a modern, end-to-end data analytics platform that enables users to conduct complex data integration, analysis, engineering and more, leveraging AI to limit the need for extensive coding. Power BI delivers AI-driven data visualization with a self-service model that allows users to chat with data, gain insights and access citations to the sources behind AI-generated answers—helping leadership develop confidence in the data and use it to make informed decisions.
Utilizing these platforms ensures scalability, best-in-class features and long-term relevance without duplicating efforts. However, this technological shift requires strong governance, including data governance and consideration of bias and ethical concerns. Companies must stay vigilant and adhere to evolving regulatory compliance to gain a competitive advantage.
“We utilize a small cross-functional, tech-savvy team that is using ChatGPT and Copilot to explore their functionality as potential tools in our overall strategic toolbox,” says Dwight. “That group is trying to understand the possibilities, risks and rewards with new tools. What we don’t want to do is get distracted by every new cool technology, so we’re trying to contain the excitement around AI to a limited group of folks.
We are also now working on an AI digitization effort beyond our factory walls, deploying and implementing full AI-driven supply chain resilience from the vendor to the factory and out to our customers. RSM is providing critical resources to assist us in this effort.
Dan Dwight, President and CEO, Cooley Group
Buie also sees the potential for AI to continue strengthening processes and decision making. “I’m focusing on empowering our executive team to talk to our data,” he says. “It sounds easy, but it’s really hard. Our goal is to create a future journey where we can have a conversation with our data—AI can provide predictive insights and then cite how it came to that insight and the process it went through. If I can enable a conversation with our data that's accurate and trusted, then I will have succeeded.”
The takeaway
Incorporating AI within your business strategy and key processes has quickly become essential, but effective implementation and integration come with challenges. As your company develops an AI strategy, success depends on several factors, including effective data governance, defined business pain points, depth of AI skills and experience, POC development, and an understanding of potential solutions. Without these elements, AI-driven outcomes may not meet expectations, leading to a loss of trust and confidence among users and stakeholders.
“Ongoing AI literacy and education will continue to remain important as technologies continue to advance,” says Beyer. “Those who have already experimented with POCs should now begin identifying where AI is delivering real value and move promising solutions into production to capture long-term, value-driven ROI. And those who are just beginning their AI journey should identify business pain points, align expectations and use platform-based foundations that can scale and stay evergreen as technologies evolve.”
AI tools and technologies may seem overwhelming at first. To drive successful outcomes and long-term value, organizations should consider working with technology providers or trusted advisors. This collaboration helps fill skill gaps, accelerate progress and support knowledge transfer so internal teams can take long-term ownership. By leveraging external guidance and advice, your organization can strategically achieve your AI goals.
Ready to get started? RSM’s experienced AI advisory team understands the enterprise AI journey and the foundational elements necessary to generate increased value and reduce risk. Contact our team to learn more about how AI can transform your key business operations.
Frequently asked questions
Several trends can help originate new models rather than just improving existing ones. For example, emerging AI solutions can create autonomous agents that independently execute common economic tasks within the business, develop hyper-personalized products and services for specific customers or groups of customers, generate continuous intelligence and observations on operations and risk, establish automated verification and compliance solutions, introduce products that improve themselves with increased velocity, and launch economies where humans and AI collaborate to jointly create value.
AI tools and applications are expected to most affect jobs by automating manual tasks, reshaping many entry‑level white‑collar positions, and significantly increasing demand for employees who can develop, manage and work closely with AI systems.
Microsoft is an effective core ecosystem, offering Fabric for data centralization and enabling integration with OpenAI. ChatGPT Enterprise and Claude are strong choices for additional use cases.
Data governance, privacy and security remain critical. To prevent data leakage or misuse, organizations must implement strict controls over where data is stored, who can access it, how models are trained and how AI tools are deployed. In addition, from a copyright and IP perspective, establishing guardrails can help prevent derivative content from being improperly reused or repurposed and limit exposure to regulatory and compliance risks.
RSM contributors
-
-
Paul SeckarPrincipal
Related insights
