top of page

The AI Advantage Is Built, Not Bought

How accounting firms can turn readiness, governance, and centralized control into safer, stronger, smarter growth


By: Christopher Stark, President & CEO, Cetrom


Artificial intelligence is quickly becoming a defining capability for accounting firms. But the firms that gain the greatest advantage will not be the ones that acquire the most tools. They will be the ones who build the discipline to use AI securely, consistently, and at scale.


For firm leaders, the question is no longer whether AI will influence audit, tax, advisory, client service, recruiting, and operations. It already does. Yet readiness lags: in an AICPA & CIMA survey of 1,446 senior finance and accounting leaders, 88% said AI will be the most transformative technology trend in the profession over the next two years, yet only 8% said their organizations are very well prepared to manage it.


The more important question, then, is whether the firm’s technology, data, policies, and people are ready for responsible adoption. Without it, experimentation creates fragmented systems, uncontrolled data movement, compliance gaps, and a new class of cyber risk. With it, AI can expand capacity, accelerate insight, strengthen client experiences, and free professionals for higher-value work.


AI Readiness Is an Enterprise Discipline

AI readiness is often treated as a software-selection exercise: choose a platform, license users, deliver training, and measure adoption. That sequence is incomplete. AI reaches into identity, access, data retention, privacy, vendor management, business continuity, ethics, and professional standards. It can surface information across repositories, generate authoritative-sounding content, and make sensitive data easier to expose. Readiness is therefore an enterprise discipline, not an isolated technology project.


A practical starting point is to assess five connected areas: business priorities, data, security, governance, and workforce capability. Which use cases will create measurable value? Where does the supporting data reside, and can it be trusted? Who should be able to access it? What rules determine acceptable use? How will employees validate outputs and escalate concerns? These questions anchor the conversation in business outcomes while exposing dependencies to resolve before broad deployment.


They also move leadership past a false choice between innovation and control, one that is already costing organizations momentum. In BDO’s 2025 Techtonic States research, 48% of executives said data security concerns are blocking AI investment. Strong controls do not slow responsible innovation; they make it repeatable. When a firm knows which data is approved, which tools are sanctioned, how access is granted, and how results are reviewed, teams can move faster with greater confidence.


GRC Must Move at the Speed of AI

Governance, risk, and compliance cannot remain a policy binder reviewed once a year. In an AI-enabled firm, GRC must become an operating system for decision-making. Governance defines who owns AI strategy, approves use cases, and sets boundaries. Risk management evaluates the potential effect on confidentiality, accuracy, independence, reputation, and client obligations. Compliance translates legal, regulatory, contractual, and professional requirements into monitorable controls.


Regulators have made the stakes explicit. In June 2026, the IRS Office of Professional Responsibility issued Alert 2026-19, its first formal guidance on how Circular 230 applies to AI. It confirms that due diligence, competence, and confidentiality duties apply in full when AI assists with the work, and that firm leadership is responsible for implementing AI policies and procedures. The AICPA’s message to members was equally clear: AI is welcome, provided it operates under human oversight, professional judgment, and strong confidentiality safeguards.


Existing obligations reach AI as well. Under the FTC Safeguards Rule, tax and accounting firms are financial institutions, regardless of size, and are required to maintain a written information security program, designate a qualified individual to oversee it, and supervise service providers. Every AI platform that touches client data belongs inside that program and the firm’s Written Information Security Plan (WISP), which IRS Publications 4557 and 5708 are designed to help firms build.


The most effective governance model is cross-functional, bringing together technology and security, risk and compliance, practice leadership, human resources, and an executive sponsor. Together, they can maintain an AI inventory, classify use cases by risk, approve vendors, define human-review requirements, and establish incident-response procedures.


Governance should also be visible to employees. Staff need clear answers: Which tools may I use? What information may I enter? When does IRC Section 7216 require client consent before tax return information is shared with an AI service? How should I verify an output? Where do I report a mistake or suspected exposure? Clarity reduces shadow AI, supports accountability, and turns every user into part of the control environment.


Centralized Data and Control Create the Foundation

AI is only as useful as the environment around it. Many firms still operate with information dispersed across local devices, legacy applications, email, file shares, cloud platforms, and third-party portals. That fragmentation obscures which version is authoritative, who can access sensitive information, and whether controls apply consistently.


Centralization does not require placing everything in one application. It means creating unified visibility and policy enforcement across the environment. Identity should be centrally managed. Access should follow role, need, and context. Data should be classified, protected, backed up, and recoverable. Approved AI services should operate within defined boundaries, and logging should provide evidence of how systems and information are used.


Visibility also determines how quickly a firm can respond to an incident. Since May 2024, the Safeguards Rule has required covered firms to notify the FTC within 30 days of discovering a security event involving unencrypted information of 500 or more consumers. A firm that cannot say where client data lives, or which AI services have touched it, will struggle to meet that clock.


This is where Zero Trust principles become essential. Every request for access should be verified rather than assumed. Privileges should be limited, and devices and identities continuously evaluated. Sensitive information should remain protected as it moves between people, applications, and AI services. These practices reduce exposure today while creating the control plane firms will need as AI becomes more embedded and autonomous.


Move from Experiments to a Managed Portfolio

Disconnected pilots rarely create enterprise value. Firms should manage AI use cases as a portfolio, prioritized by value, feasibility, and risk. A focused roadmap might begin with lower-risk internal applications such as knowledge discovery, meeting summaries, drafting support, or administrative automation. Use cases involving client data, professional judgment, or externally delivered work require stronger validation, supervision, and documentation.


Each initiative should have an accountable owner, a defined problem, approved data sources, success measures, and a review date. Useful measures include hours returned to professionals, cycle-time reduction, adoption by intended users, output accuracy, policy exceptions, and security events. If a use case cannot demonstrate value or stay within the firm’s risk tolerance, it should be redesigned, contained, or retired.


Training must evolve as well. Professionals need role-based education that combines practical workflows with data-handling rules, critical thinking, output verification, and escalation procedures. Leaders should reinforce what the IRS guidance now makes explicit: AI can assist judgment, but it does not replace professional accountability.


Partnering with Experts Is a Strategic Multiplier

Accounting firms do not need to navigate this transformation alone. The challenge is not a shortage of AI products; it is connecting strategy, infrastructure, cybersecurity, compliance, and change management. An experienced technology and security partner can help leaders assess current-state maturity, identify hidden dependencies, establish a practical governance model, and sequence investments according to business priorities.


The right partner understands accounting, sensitive data, busy season, and IRS and FTC oversight. It unifies managed IT, cybersecurity, compliance, AI governance, and strategic advisory, while challenging assumptions and translating technical risk into executive decisions.


Accounting firms move fastest when leaders own strategy and judgment, while a partner provides expertise, continuous monitoring, and a path from assessment to execution. This creates a durable model for evaluating each tool and threat.


Safer. Stronger. Smarter.

The next phase of AI will reward firms that prepare before pressure forces action. Those firms will know where their data lives, how it is protected, which tools are approved, who is accountable, and how value is measured. They will treat governance as an enabler, not an obstacle, and build with partners who understand both the opportunity and the stakes.


Safer, because data, access, and risk are governed. Stronger, because technology and people operate from a resilient foundation. Smarter, because AI investment is tied to measurable business outcomes.


The competitive advantage will not come from being first to every new capability. It will come from being ready to adopt the right capabilities responsibly and to scale them with confidence. For accounting firms, that is how AI becomes more than an experiment. It becomes a platform for trusted growth.


About Cetrom: Cetrom is an accounting-focused managed IT and cybersecurity provider (MSP and MSSP) that helps firms modernize securely, strengthen compliance, and adopt AI responsibly. Through Zero Trust architecture, strategic IT advisory services, and white-glove support, Cetrom gives firm leaders the control and confidence to turn technology into a platform for resilient growth. Assess your firm’s AI readiness at cetrom.com.


Sources

AICPA & CIMA, “AI Transformation Opens Door for Finance Professionals to Build Future- Ready Skills” (Future-Ready Finance: Technology, Productivity, and Skills Survey), December 17, 2025.

AICPA & CIMA, “AICPA Insights: Introductory Guidelines for Responsible AI Use in Federal Tax Practice,” July 2, 2026.

BDO USA, 2025 Techtonic States Report, Chapter 3: Secure Your Business Edge, 2025.

IRS Office of Professional Responsibility, Alert 2026-19, Introductory Guidelines for Responsible AI Use in Federal Tax Practice, June 24, 2026.

IRS Publication 4557, Safeguarding Taxpayer Data, and IRS Publication 5708, Creating a Written Information Security Plan for Your Tax & Accounting Practice.

Internal Revenue Code Section 7216 and Treas. Reg. §§ 301.7216-1 through 301.7216-3 (disclosure and use of tax return information by preparers).

Federal Trade Commission, Standards for Safeguarding Customer Information (Safeguards Rule), 16 CFR Part 314, including the notification amendment effective May 13, 2024.

Comments


bottom of page