AI Adoption in Accounting Firms: The Risks and Opportunities for MSPs

AI could change how accounting firms work. Without the right safeguards, faster output could create bigger risks.

AI could help accounting firms serve more clients, but it also raises questions about data privacy, cybersecurity and accountability.

In the Is Tech Driving the Future Firm? MasterClass, Jason Croston, Managing Partner at SRJ Walker Wayland, describes both sides of this challenge. He sees AI as a way to expand what accounting professionals can deliver, yet worries about employees sharing inappropriate data with AI tools and the quality of AI-generated output.

His perspective reveals a challenge for managed service providers: helping accounting firms explore AI without treating access to a tool as proof that the firm is ready to use it.

Will AI Replace Accountants or Change What They Can Deliver?

The question of whether AI will replace accountants often dominates the conversation. Jason takes a different view.

He explains that accounting professionals already struggle to find enough time to do everything they would like for their clients. AI could create capacity for more client conversations, tax planning and work that might otherwise go unfinished.

That makes AI in accounting more than a discussion about completing routine tasks faster. Its potential also lies in giving professionals more time to apply their expertise where clients need it.

AI for accountants can create value. But firms must consider where human judgement remains essential and how employees will use these tools in their work.

Jason challenges the assumption that greater efficiency necessarily leads to fewer people being needed. The opportunity may instead be to use available capacity differently and deliver more value to clients.

For MSPs, helping a client explore AI’s potential is one conversation; addressing the risks that come with using it is another.

Shadow AI: The Risk of Use You Can’t See

An accounting firm may introduce approved AI tools and establish usage rules, yet still struggle to know how employees use them in practice.

This is known as shadow AI: employees using AI tools outside the organisation’s approved processes or oversight. For example, an employee might paste a client document into a public chatbot to summarise it, unintentionally exposing confidential information.

Jason raises this concern directly. Even with a written AI policy, management can struggle to know whether employees are following it.

For MSPs, the challenge is helping firms close that gap. They can start by asking:

• Which AI tools are employees using?
• Are those tools approved for the intended tasks?
• Do employees understand what information they must not enter?
• Is there a clear process for raising questions about unfamiliar tools or uses?

The answer is not to prohibit experimentation. It’s to reduce unnecessary exposure and give employees a clear, workable way to use AI within the firm’s requirements.

An AI Governance Framework Must Work Beyond Paper

Identifying shadow AI is only one part of the challenge. Firms also need a practical approach to managing approved AI use.

An AI governance framework establishes the rules, responsibilities and controls that guide how AI is introduced and used. For accounting firms, it needs to account for sensitive client information, the risks of inaccurate output and the professional responsibilities employees retain when using AI.

A written policy is a starting point. But it doesn’t guarantee consistent behaviour. Employees need clear guidance, managers need appropriate oversight and the firm needs a way to address problems when they arise.

MSPs can help clients assess the technical and operational controls supporting that framework, including:

• Access and permissions: Who can use AI tools and what information can they access?
• Data handling: What restrictions apply to entering client information or confidential documents into AI tools?
• Approved use: Which applications are permitted and where should AI use be restricted?
• Oversight and accountability: Who reviews AI use and responds to incidents?
• Employee guidance: Do staff understand how the rules apply to their everyday work?

These considerations are not a complete compliance checklist. The controls required will depend on the firm’s systems, obligations and intended uses of AI.

Jason’s point is clear: guardrails should manage risk whilst giving employees the confidence to experiment with AI.

AI Output Still Needs Quality Control

Data security is only part of Jason’s concern. He also points to the risk of poor-quality AI output being accepted without adequate checking.

AI-generated content can sound convincing while containing errors or omissions. In accounting, polished output is not the same as reliable work.

Firms need to determine which outputs require verification, who checks them and where AI assistance is inappropriate. For instance, an AI-generated first draft is not equivalent to verified financial information or work that affects a client’s obligations.

MSPs can help clients understand the technology’s capabilities and limitations and support suitable technical safeguards. Accounting professionals must retain responsibility for relevant professional judgements and approvals.

When considering how to use AI in accounting, firms must look beyond whether a tool produces a useful result to whether that result is fit for its intended purpose.

Copilot Readiness: Is the Environment Ready for AI?

For firms exploring Microsoft Copilot or other AI-enabled workplace tools, readiness involves more than making a product available.

A Copilot readiness assessment can help identify whether the firm’s environment and working practices are prepared for the intended use. Depending on the product and scope, this might involve reviewing permissions, access to information, data exposure, security settings and employee readiness.

AI tools can surface information users already have permission to access. If permissions or information-management practices are poorly configured, introducing AI can expose existing weaknesses.

Before deployment, MSPs can help accounting firms assess:

• Access permissions: Ensure employees can access only information appropriate to their roles
• Data protection: Protect sensitive information
• Employee readiness: Prepare staff to understand the tool’s intended uses and limitations
• Policies and controls: Establish appropriate safeguards and support processes

The assessment should match the technology being introduced. Readiness for one Copilot deployment doesn’t establish that every AI application is safe or appropriate.

For MSPs, the value lies in preparing clients for a specific deployment, not treating AI adoption as a simple software activation. For firms already using Microsoft, getting more from Microsoft 365 is a related consideration, particularly when evaluating Copilot and the tools employees already have available.

What Selling AI Services Means for MSPs

As interest in AI grows, MSPs can see opportunities in selling AI services to accounting-firm clients. But the conversation shouldn’t begin and end with access to a product.

Clients may need help with deployment, security, appropriate use and the controls that support ongoing adoption. MSPs must be clear about what they can support, where their responsibilities end and when specialist advice is needed.

MSPs offering AI for MSPs services must address specific client needs and set realistic expectations for productivity gains and risk reduction.

Before recommending a solution, MSPs need to understand the accounting firm’s actual requirements and risks. As explored in how MSPs can help accounting firms make smarter technology decisions, technology investments should serve a clear business need rather than follow the appeal of a new capability.

AI services will be more credible when they address a defined need and set realistic expectations about the support provided.

The Bigger Question: How Can Accounting Firms Experiment With Confidence?

AI adoption in accounting firms is not simply about introducing new tools. It’s about unlocking their value without compromising sensitive data, work quality or accountability.

Jason highlights the challenge: AI could give accounting professionals more capacity to serve clients, but data privacy and unreliable output remain concerns.

MSPs can help firms adopt AI responsibly by preparing their environments, implementing appropriate safeguards and understanding the risks of each deployment.

What does an accounting-firm leader need from technology providers as AI reshapes the profession?

In the Is Tech Driving the Future Firm? MasterClass, Jason shares his perspective on AI’s potential, the risks that concern him and what MSPs need to understand about the firms they support.

Get Access to the MasterClass