AI Compliance in Accounting: A Practical Guide for Internal Policies

An AI policy for an accounting firm sets out which AI tools staff may use, what client data those tools can touch, who reviews AI output before it reaches a client or the IRS, and how the firm checks that the rules are followed. To build one, list the AI already in use, draft the rules with IT and legal, train the team, then audit and update the policy as tools and regulations change.

You're looking to integrate AI into your CPA firm while staying compliant and efficient? This guide walks you through the essential steps to set AI policies tailored for accountants. From ethical use and data privacy to training your team and monitoring compliance, you'll learn how to harness AI's potential while protecting your firm and clients.

As an accountant in the US, you already know the grind, especially during tax season.

With the rise of artificial intelligence (AI), it's time to start thinking about how you can use this technology to lighten your workload, boost accuracy, and stay ahead of the competition. But jumping into AI without a plan can be risky.

That’s why setting clear AI policies is crucial.

In this guide, we’ll walk you through the steps to establish AI policies that fit the unique needs of your firm.

Why do US accountants need AI policies?

A written AI policy keeps client data protected, keeps a professional accountable for every AI-assisted return, and gives staff clear rules on when to trust the tool. We all know AI is transforming the way we work, but what does that mean for you as an accountant? Well, AI can handle the repetitive, time-consuming tasks like data entry, document review, and number crunching, giving you more time to focus on what really matters: strategic decision-making and client relationships.

However, jumping into AI without a plan isn’t smart. That's where AI policies come in.

These policies help you ensure that AI is being used ethically, that data privacy is protected, and that your team is trained properly. In short, it’s about making sure AI works for you, not the other way around.

Some key concerns for accountants in the US include:

• Ethical Use: How do you make sure AI decisions are fair and unbiased?
• Compliance: With IRS regulations, how do you ensure AI stays on the right side of the law?
• Training: Your team needs to know how to use AI effectively without relying on it too much.

What should a CPA firm's AI policy include?

At minimum, a CPA firm's AI policy should cover ethical use, data privacy and security, and governance (who oversees the tools and enforces the rules). To ensure AI enhances your firm’s operations, these are the key elements that should be part of your AI policy:

1. Ethical use of AI
The accounting industry revolves around trust and fairness. Here’s how you can uphold these principles with AI:

• Transparency: Make AI processes clear and understandable to your team and clients.
• Fairness: Avoid biases in algorithms that could impact financial decisions.
• Accountability: Define who's responsible for AI-generated outcomes, whether it’s a wrong tax filing or missed deduction.

2. Data privacy and security
With sensitive financial data at stake, keeping it secure is top priority. Here’s how:

• Data Protection: Use advanced encryption to safeguard client information.
• Compliance with IRS Guidelines: Ensure AI tools comply with US regulations like the IRS’s e-file security standards (set out in IRS Publication 1345) and the FTC Safeguards Rule, which applies to tax preparers.
• Access Controls: Limit data access to trusted employees only, reducing the chance of a data breach.

3. AI governance
Accountants thrive on structure, and your AI policies should reflect that. Put people in charge of overseeing AI tools and ensuring they’re being used correctly:

• Leadership: Assign a team or committee to manage AI implementation and oversight.
• Monitoring: Regularly audit your AI tools to check for accuracy and compliance. The NIST AI Risk Management Framework is a useful free reference for what to check.
• Policy Enforcement: Set up clear rules for addressing any misuse of AI.

How do you implement an AI policy in your firm?

Implementation takes four steps: take stock of current AI use, draft the policy, communicate and train, then monitor and adjust. Now that we’ve covered why AI policies are important, let’s get into the “how.” Here are some practical steps to put these policies in place at your firm:

Step 1: Take stock of your current AI usage
Before creating new policies, figure out where AI is already being used in your firm. Ask yourself:

• What AI tools are we using? Are you using AI-driven tax software? Or perhaps AI for client communications? (This overview of AI tools for tax professionals can help you map the categories.)
• How well are these tools working? Do they meet your needs, or are there gaps?

Step 2: Draft your AI policy
Now that you know where you stand, draft clear guidelines for AI use:

• Set Clear Expectations: Outline how your team should (and shouldn’t) use AI. For example, when is it appropriate to rely on AI vs. human judgment?
• Collaborate with Experts: Involve your legal, IT, and HR teams to make sure your policy covers all bases.
• Stay Aligned with Core Values: Ensure your AI policy reflects the ethical standards that make your firm trustworthy.

Step 3: Communicate and train
Once the policy is ready, it’s time to roll it out across your firm:

• Training Sessions: Teach your team how to use AI tools and explain the policy in detail.
• Accessible Guidelines: Keep the policy easy to access, whether it’s on your firm’s intranet or via printed handouts.
• Ongoing Feedback: Create a space for employees to ask questions and offer feedback on AI usage.

Step 4: Monitor and adjust
AI is always evolving, and so should your policy:

• Regular Audits: Conduct periodic checks to ensure your AI tools are following the policy.
• Feedback Loops: Encourage employees to report issues or suggest improvements.
• Policy Updates: Be ready to tweak the policy as new AI capabilities emerge or regulations change.

Where does Filed fit into your firm's AI policy?

Filed is the kind of tool your policy should cover: it prepares and reviews returns, traces each number back to its source document, and leaves sign-off with your reviewers.

You’re busy, especially when tax season hits.

That’s where Filed comes in.

This AI assistant can save you hours by automating tedious tasks like data entry and error-checking.

Here’s what Filed brings to the table:

• Automated Data Entry: Filed Prep reads client documents, including bank statements and scanned docs, and prepares the return in your tax software. Say goodbye to manual entry and human error.
• Real-Time Error Detection: AI cross-checking in Filed Reviewer helps catch mistakes before they become a problem, supporting compliance and accuracy. Filed scored 94% line-by-line accuracy on TaxCalcBench.
• Direct Integration: Whether you use CCH Axcess, Lacerte, Drake, or ProConnect, Filed works with your tax software.
• Cut Costs: By automating routine work, Filed can cut your tax prep costs by up to 50%. That’s more money in your pocket or more time for high-value tasks.
• Top-Level Security: Filed is SOC 2 compliant and ISO 27001 certified, and uses encryption and access controls to keep your client data safe.

Sound too good to be true?

Book a demo and see for yourself how AI can transform your tax season!

Frequently asked questions

What should an AI policy for an accounting firm cover?

Which AI tools staff may use, what client data they can access, who reviews AI output before it goes to a client or the IRS, how staff are trained, and how often the firm audits compliance with the policy.

Do data security rules apply to AI tools that handle client tax data?

Yes. The same obligations apply whether a person or an AI tool handles the data. Tax preparers fall under the FTC Safeguards Rule, which requires a written information security program, and federal law limits how preparers use and disclose return information.

Who is responsible when an AI tool gets a return wrong?

The professional who signs the return. AI can prepare and check the work, but accountability stays with the preparer and the firm, which is why your policy should require human review of AI output.

Is Filed secure enough to fit a firm's AI policy?

Filed is SOC 2 compliant and ISO 27001 certified, and SOC 2 and ISO reports are available under NDA. Access is permissioned through your firm's Filed workspace, and sign-off on every return stays with your reviewers. Details are on the Filed trust center.

Conclusion

AI is here to stay, and US accountants who adopt it wisely will gain a competitive edge.

By setting clear, ethical AI policies, you can protect your firm, your clients, and your team.

Not only will you ensure compliance, but you’ll also maximise the benefits AI has to offer, like boosting efficiency, improving accuracy, and cutting costs.

Ready to make the leap?

Filed's AI tools are designed specifically for accountants like you.

Try it out and change the way you work this tax season!