AI skeptic to AI ready

The changing face of tax in 2025

Pixelated cursor hand touching a classical painting hand.
Logo of Filed

This report draws insights from both sides: tax pros who live in the trenches (you know, the ones answering emails at 2am in busy season, surviving on their fifth cup of coffee) and the technical innovators building tomorrow's tools.

We examine what AI can do today, where it falls short, and the 5 forces reshaping tax in the AI era.

Consider this your practical guide to navigating tax's technological shift, written for skeptics and believers alike.

The AI tax landscape by the numbers

Businesspeople walking into a giant suitcase.
Security’s price tag
$4.8 million
the cost of an average data breach
The talent drain is real
300,000+
accountants have left the profession in recent years, while the number of new CPA exam takers has plummeted by 34%
Man in a suit looking to the side.
The economics of efficiency
50%
reduction in processing costs, plus a 30% boost in capacity with AI-powered tax prep
AI’s edge
2/3
of accounting professionals now see AI as key to staying ahead

The tax pro’s tech dictionary

AI

n. Artificial Intelligence

Software that mimics human thinking to solve problems, recognize patterns, and make decisions. Like having a super efficient assistant who never gets tired.

SOC2

n. System & Organization Controls 2

A rigorous compliance standard that outlines criteria for managing customer data based on five principles: security, availability, processing integrity, confidentiality, and privacy.

On-prem

n. On-premises

AI software that runs on your firm's local servers rather than being hosted in the cloud.

RPA

n. Robotic Process Automation

Software that automates repetitive tasks, like data entry or moving files around, without human intervention.

AI Agents

n.

Advanced, autonomous software systems that perform specific tasks using AI, like identifying inconsistencies in tax data or summarizing financial reports.

Cloud Technologies

n.

Storing and processing data over the internet rather than on your office servers. It’s convenient and flexible, but you’ll want to make sure the security is solid before you trust it with client data.

OCR

n. Optical Character Recognition

Tech that reads and converts printed or handwritten text into digital data. It’s how your tax software pulls numbers from client documents without you squinting at PDFs for hours.

The AI inflection point

A man labeled "Skeptic" sits at a desk with papers, surrounded by symbols of technology, AI, and data, looking doubtful.
01
The math doesn't work anymore for traditional tax prep.

With over 300,000 accountants leaving their jobs in recent years and CPA exam takers down by 34%, firms processing 250+ returns face a reality they can’t ignore: there simply aren't enough hands to do the work the old way. Something has to give.

02
Offshore outsourcing—the last decade's efficiency play—is showing its cracks.

While outsourcing overseas addresses immediate capacity issues, it comes with baggage: sensitive client data crosses borders, the 7216 consent form that raises clients’ eyebrows, time zone gymnastics that delay turnaround, and quality control that eats into promised savings.

What once seemed like a smart cost-cutting measure is now a balancing act between savings and expenses.

03
Meanwhile, machine learning has (not so) quietly crossed the threshold from promising to practical.

The advances aren't theoretical anymore; they're operational. Today's AI can interpret complex tax scenarios, apply contextual reasoning, and execute prep tasks with a precision that rivals professional work.

04
The smart firms aren't waiting for permission to evolve.

Early adopters are already reporting a 50-90% reduction in data entry time, with the average firm reclaiming 4-6 hours per week per staff member.

With 66% of accounting professionals now recognizing AI as a competitive advantage, the conversation has shifted from ‘if’ to ‘how’.

05
Yet skepticism persists for good reason.

In a profession governed by strict regulation and fiduciary responsibilities, caution with client data isn't just prudent… it's non-negotiable.

Over a third of firms rank data privacy as their primary concern, and they're asking the right questions.

06
AI represents a significant shift in how tax firms operate.

The stakes? High.

The potential rewards? Substantial.

The need for careful decision-making? Greater than ever.

The 5 forces reshaping tax
in the AI era

AI is already separating firms into two camps: those using it to reimagine how they work and those hoping it’ll go away.

The firms winning see AI not just as software but as a strategic lens for reimagining professional practice. The firms losing? They’re clutching their spreadsheets, waiting for a sign from the IRS that it’s time to adapt.

[ AI isn’t for every CPA–yet. ]

Smaller firms might jump in too early, paying for AI tools they aren’t positioned to maximize.

Specialized practices working with highly complex clients might still need more human touch than today’s AI can offer.

Legacy firms already running like clockwork with loyal staff might not see enough immediate upside to justify the change.

The key is knowing where your firm stands and making deliberate, strategic choices about when and how to use AI.

‘CPAs often focus on the wrong thing with AI. It’s not about replacing people — it’s about removing grunt work so they can focus on high-margin advisory. The real shift happens when CPAs think strategically, not just technically.’

Mohson Choudhry avatar
Mohson Choudhry
CPA, Ayub Accounting
AI isn’t a trend; it’s a business decision

Through our research and conversations with CPAs across the country, we’ve identified 5 major forces reshaping tax work in the AI area: control over data, evolving economics, quality assurance, technology without turbulence, and the reinvention of the tax pro.

These forces might feel like abstract ideas, but they’re the practical guardrails protecting a CPAs most valued assets: client trust, professional reputation, and economic sustainability.

At their core, these forces cut to the heart of CPAs’ most pressing concerns:

01
How do I protect my clients’ sensitive information?
02
What’s the advantage of using AI agents over offshore staff?
03
Can AI really handle complex tax scenarios?
04
Will this improve or completely disrupt how I work?
05
What happens to my professional value with AI?
Force #1

Control over data

Multiple security cameras mounted on a pole.
68%
of tax professionals rank data security among their top concerns when evaluating AI tools

When a client hands over their W2s and 1040 returns, they're not just offloading paperwork; they're trusting you with their professional and personal vulnerability.

But you already know that.

That’s why security concerns are still the number one roadblock to AI adoption among tax pros. A whopping 68% of tax professionals rank data security among their top concerns when evaluating AI tools.

15% of accounting firms have experienced a data breach, and the average cost of one has rocketed to $4.8 million. The fallout isn’t just expensive; it’s enough to end most businesses.

Remember when BDO USA had to cough up $10 million in a Federal Trade Commission settlement? The Big 4 can absorb a financial hit like that. Most firms can’t, and the cost isn’t just in fines. It’s in client trust and reputation.

The security paradox 
of growth

While no CPA wants to make their data vulnerable, scaling your practice inevitably introduces new security variables.

Every time you extend your practice's capacity—whether by hiring seasonal staff, contracting with offshore teams, or rolling out new tech—you create more risk entry points. Even firms that aren’t actively growing but use offshore processing to cut costs can’t always say, with confidence, where their clients' data actually lives.

And if you’re sending client tax documents to Vietnam or India via Google Drive, let’s be honest: you’re just hoping the security is up to standard.

Different business models handle these challenges in different ways.

The table below breaks down the core differences and what they mean for your firm.

Offshore
International data centres
high risk
Limited visibility into security protocols
Inhouse
Local servers or software
medium risk
Requires significant manual handling
AI
US-based servers
low-medium risk
Scales easily without compliance headaches
Gavel next to a document with the number 7216.

In practice, these models create dramatically different client data journeys.

Traditional processing means a W-2 passes through four or five hands, each one a risk. Even solid internal protocols crack under tax season pressure.

Offshore outsourcing cuts costs but sends data across borders, relying on third-party security practices. Once it’s out there, control is gone.

With AI processing, data stays on domestic servers, encrypted, with fewer human touch points. But not all AI solutions are built equally—some scatter data across multiple servers.

For CPAs thinking ahead, security is evolving beyond checking compliance boxes.

It's becoming a strategic differentiator in an industry where client trust remains the foundation of success.

‘7216 causes discomfort. Clients expect their financial info to stay close.’

Leroy Kerry avatar
Leroy Kerry
CEO, Filed

Questions that keep tax pros up at night

Tax professional surrounded by hands holding question cards.
01
Who sees this data?

With traditional processing, it's your staff. With offshore outsourcing, it's whoever is working on it overseas. With AI, fewer people have access, usually limited to specific personnel with authorized privileges.

02
Where does it live?

Client data may reside on local servers, international data centers, or encrypted cloud environments. The location determines which legal jurisdictions apply, what protections exist, and what options you have if there’s a breach.

03
Can it be permanently deleted?

With traditional processing, access is as strong as your internal policies and office door. Offshore outsourcing? It’s in someone else’s hands.  AI systems mean fewer people handle data, access is tracked, and multi-factor authentication is standard.

04
The smart firms aren't waiting for permission to evolve.

Many clients assume their data can be erased when they request it. But in traditional or offshore models, that’s not always clear. Modern AI tools, like Filed, offer verifiable deletion protocols that can be executed on demand.

Evaluating AI security is different because you need an understanding of technical safeguards and human oversight. Traditional methods let you trace who touches client documents. With AI, CPAs should be asking which parts of processing are fully automated, which are human-reviewed, and what security controls protect each stage.

‘Filed uses the same encryption that protects CPAs’ bank accounts. What was once enterprise-grade is now standard.’

Atul Ramachandran avatar
Atul Ramachandran
CTO, Filed

AI enhances security, but it won’t take the fall for you

AI-powered tax prep tools improve efficiency and introduce more structured security controls. Fewer handoffs, consistent protocols, and data encryption reduce risk compared to traditional and offshore models. But no system is airtight. If something goes wrong, liability still falls on the CPA.

[ Robust review processes are non-negotiable. ]

A poorly implemented AI solution can create just as many security risks as it solves—new vulnerabilities, new ways for bad actors to weasel in, and plenty of fresh compliance nightmares if you don’t ask the right questions upfront. No matter how advanced it is, AI should be seen as an assistive tool, not a decision-maker.

In a profession built on precision, chances are you’re also wondering: If AI can safeguard my data, can it also protect my bottom line?

Integrate AI without legal risks

SOC2 Compliance
Your provider meets top-tier security standards verified by independent auditors.
Bank-Grade Encryption
AES-256 and 256-bit SSL keep your data locked down, whether it’s moving or stored.
Onshore Data Processing
Keep your data within the US, making compliance with domestic regulations simple.
Automated Compliance Alerts
Your AI will spot potential regulatory issues early before they become problems.
Reputable Cloud Partners
Pick infrastructure hosted by trusted providers like Microsoft with enterprise-grade security.
Regular Algorithm Updates
Keeps AI in sync with changing tax laws so you don’t have to.

Evaluating AI security is different because you need an understanding of technical safeguards and human oversight. Traditional methods let you trace who touches client documents. With AI, CPAs should be asking which parts of processing are fully automated, which are human-reviewed, and what security controls protect each stage.

‘The smartest firms aren’t adopting AI because it’s trendy. They look into what protects client data and scales efficiently.’

Leroy Kerry avatar
Leroy Kerry
CEO, Filed
Force #2

Evolving economics

Layers of torn paper resembling evolving.
83%
of IT leaders turn to offshore solutions, especially in peak season.

Offshore outsourcing has been the 
go-to move for firms looking to cut costs for years.

On paper, it’s a no-brainer: lower costs, more hands on deck, and an easy way to scale. But tax pros don’t just work on paper. They work in the real world, where efficiency isn’t just about who does the work but how well it gets done.

Firms that rely on offshore teams to immediately slash overheads or scale quickly often run into day-to-day friction that eats into productivity:

Communication barriers turn simple requests into endless email chains.
Quality control lapses require extra oversight, negating promised savings.
Compliance headaches arise when client data crosses borders–try explaining a 7216 consent form to a skeptical client.

Offshore vs onshore: a shift in how firms think about value

Tech doesn’t just make tax prep faster; it’s forcing CPAs to rethink what efficiency means.

Offshore outsourcing is all about labor arbitrage: shifting work to lower-wage regions. AI offers something different: intelligent resource allocation. Instead of shifting the grunt work elsewhere, AI eliminates it altogether. And it doesn’t have to be an all-or-nothing choice.

Many larger firms won’t abandon offshore teams overnight. Instead, early adopters are blending AI with offshore models: AI handles data-heavy prep work, offshore teams provide scalable human review, and in-house pros focus on client strategy.

Essentially, AI handles the grind so tax pros don’t have to. Here’s how it makes a difference:

  • Keeps sensitive data onshore, eliminating cross-border security risks and tricky 7216 consent form issues.
  • Reduces errors, cutting back-and-forth corrections and late night fixes.
  • Remembers your clients and your workflow, so you’re not context switching, chasing last year’s figures or training new staff on things they’ll forget next season.
  • Lets you scale without adding headcount. Because who has time to train another temp?

And when you break down the real costs of offshore vs. AI, the savings don’t always land where you expect. Industry research suggests firms switching to AI can save up to $50-100 per return, and not just on headcount.

‘Firms often want to turn their low-cost center into a profit center. But if you've got a team in the Philippines producing work, and your US team has to spend time fixing it, they are wasting their time.’

Leroy Kerry avatar
Leroy Kerry
CEO, Filed

Where the savings show up

Hand holding a stopwatch.
Tax prep time shrinks by 50-90%
AI automates the repetitive work, so your team isn’t stuck on data entry.
4-6 hrs saved weekly per staff
Less time on prep means more time for high-value work: client relationships, tax strategy, and advisory services.
Man in a suit smiling and looking down.
Two tax professionals having a conversation over coffee.
$0 in infrastructure costs
No need for extra office space, IT setup, or contracts with offshore providers.
No training, no onboarding, 
no handovers
Offshore staff come and go. AI doesn’t need an orientation.
Lower labor costs
AI scales up without the stress of recruiting, training, and managing more people.
Man in a suit and hat sitting and looking through binoculars.

Offshore outsourcing spreads costs across headcount—more people, more wages, more oversight.

AI, on the other hand, is a fixed investment: licensing fees, computing power, and workflow integration.

Long-term savings depend on how well firms integrate AI into their processes. For high-volume firms, AI pays off fast. But if your firm only handles a handful of returns, those savings won’t show up overnight.

The hidden costs that can add up:

1. AI oversight: AI models don’t quit but do need monitoring to maintain accuracy and compliance.

2. Offshore inefficiencies: Time zone lags and language barriers mean those savings start looking smaller.

3. Hybrid model: Upfront costs may rise before long-term savings take hold.

‘AI like Filed is ~10x cheaper than offshore.
As costs drop, CPAs will shift from prep work to high-value advisory that builds meaningful client relationships.’

Kevin Lee avatar
Kevin Lee
CPA, Filed

So, what’s
the smarter investment?

Neither choice is inherently wrong. But the firms tracking where their hours (and dollars) actually go—not just what they spend upfront—will come out ahead.

As data entry and standard returns get automated, premium pricing will be harder to justify. Firms that cling to old workflows or stick to offshore as a forever fix will find themselves competing on price, not expertise.

And that’s a race to the bottom.

[ There’s no denying tax prep is getting commoditized. ]

At the end of the day, clients don’t pay for tax prep. They pay for tax pros who know how to think.

People typing on old computers with torn faces revealing paper underneath.

‘You shouldn't worry about AI taking your job, but you might worry about a tech-enabled CPA taking your customer.’

Leroy Kerry avatar
Leroy Kerry
CEO, Filed
Force #3

Quality assurance

Hand holding a magnifying glass.

Tax pros don’t get the luxury of being ‘close enough’.

A single mistake can trigger an audit, rack up penalties, and shake client trust.

So, while most CPAs can see how AI-assisted tax prep introduces new efficiencies, skeptics are still wondering whether it can deliver the same level of precision and professional rigor they demand.

‘When you prep a return yourself, you know what you need and have a sense of the client’s profile. But when AI does it, you’re flying blind. That’s where the skepticism comes from: practitioners might worry about what's being entered, where, and why.’

Kevin Lee avatar
Kevin Lee
CPA, Filed

A system built for consistency

AI doesn’t have off days.It doesn’t get distracted, rush through returns at 11pm, or let attention slip in the middle of a brutal tax season. Where human accuracy fluctuates, AI applies structured, repeatable logic across every return.

Think about it:

Human experience levels vary
AI doesn’t. A junior tax professional might interpret tax law differently from a partner. AI standardizes calculations  and applies tax rules the same way every time.
Even with experience, mistakes happen
AI flags them before they do. The best AI tax prep tools don’t just process returns; they proactively catch errors, spot inconsistencies, and highlight compliance risks before a CPA begins their review.
AI benchmarks now show near-perfect accuracy
Recent testing shows advanced AI models can now pass portions of the CPA exam—imagine what specialized tax.

AI like Filed can achieve when it's  built specifically for the work you do every day.

But accuracy is only part of the equation. In tax prep, precision is also about ensuring the wrong answers never get through.

Think of AI as the tax pro who never sleeps or second-guesses. Instead of relying on review cycles to catch errors, it’s built to prevent them (and make compliance easier).

So, how does AI tax prep quality assurance work in the real world?

01
Learning all the time

A CPA flags an issue on a return processed by an AI tool. The AI doesn’t just move on; it remembers. It refines its accuracy, integrating real-world tax data and CPA feedback so that the same issue doesn’t show up again next year.

02
Catching errors before they matter

A return is missing a key document. Before the CPA even notices, AI has sent them an alert. Intelligent error detection means fewer post-filing headaches, fewer awkward client calls, and fewer nights spent fixing what should have been caught upfront.

03
Transparency over trust-me math

An AI tax tool spits out a result. You want to know why. The best AI systems show their workings. Every deduction, exception, and flagged risk comes with a clear, audit-ready trail of why that decision was made.

‘Typically with AI you type something in and you get magical outputs back. That doesn't really work for CPAs; they want to know where every number came from. With Filed, we’ve built prep notes that clearly trace each value.’

Atul Ramachandran avatar
Atul Ramachandran
CTO, Filed

Don’t lose the human touch

At the moment AI thrives in rule-based scenarios: think W-2 filers, standard deductions, and well-documented business returns. When everything follows a pattern, AI executes with near-perfect accuracy.

But you know better than anyone: tax isn’t always neat.

A multi-entity business restructuring mid-year? AI might hesitate when applying tax treatment across state lines.

A high-net-worth client with layered income streams, trusts, and overseas holdings? AI isn’t making the judgment calls on tax positions; that’s still your job.

A gray area in tax law that requires professional discretion? AI doesn’t interpret nuance. It flags, and CPAs decide.

Where CPAs are essential
Navigating gray area deductions and judgment calls
Intricate multi-entity tax strategies
Understanding client nuances and intent
Where AI Excels
Spotting missing documents before you do
Surfacing IRS compliance risks early
Catching data entry slip-ups

‘AI still struggles with complex, multi-entity returns. But in five years, we'll see a complete transformation of where machine learning can add value.’

Leroy Kerry avatar
Leroy Kerry
CEO, Filed

So, where does this leave us?

The real question isn't whether AI can match human accuracy (it already does for standard returns, and unlike your seasonal intern, it doesn't need constant caffeine to maintain performance).

It's: how can you use algorithms to handle the mind-numbing data entry while you focus on the parts of tax practice where nuance, experience, and client relationships make all the difference?

Even with quality concerns addressed, there might still be that nagging voice in your head asking, ‘Do I need to rewire my entire office and learn to code just to implement this thing? Or can I actually use it without earning another degree?’

Force #4

Technology without turbulence

Man in a suit looking through binoculars and smiling.

You've spent years perfecting your firm's processes. The last thing you need is some shiny new tech that requires you to tear up your playbook and start from scratch.

Let's address the elephant in the room:

[ How much disruption are we really talking about with AI tax prep? ]

Invisible tech, visible results

AI in tax prep doesn’t work in a vacuum…it must fit into a firm’s existing workflows, software stack, and client service model without creating chaos. It shouldn’t feel like another system to babysit or a new platform to memorize.

The real magic happens when it amplifies what you’re already doing, getting the job done without making a fuss.

The best AI agents eliminate friction in tax prep without disrupting the way CPAs already work.

Think of AI-powered tax prep as adding a turbocharger to your existing engine, not replacing the whole car. Tools like Filed work directly within your existing software environment—there’s no new platform to learn.

Using Robotic Process Automation, the AI pulls client data from your existing storage systems and uploads review-ready returns into your tax software (think Drake, ProConnect, CCH Axcess and more).

It’s designed this way because tax pros shouldn’t have to choose between familiar software and reclaiming their time.

‘Our AI reads and categorizes client documents, figuring out what’s needed for your tax software. This entire process is automated behind the scenes, and it then enters data directly into your tax software so you don't have to.’

Atul Ramachandran avatar
Atul Ramachandran
CTO, Filed

Getting
your team on board

So, the tech fits right into your workflow. Problem solved, right? 
Not quite.

Most firms find their biggest implementation challenge with AI isn't the software. It's Bob who's been doing returns the same way since the Clinton administration.

What really throws a wrench in the works isn't the tech; it's the humans using it. Your staff might secretly wonder: ‘Is this going to make my job harder? Am I about to become AI's babysitter instead of a tax professional?

’Ironically, the same people who threatened to quit during the busy season might now defend manual data entry like it's a sacred art form. Funny how that works.

This resistance isn’t uncommon. Despite the hype, AI adoption across all industries is in its early days. Nationally, 7% of businesses are using AI right now, with projections showing that number creeping up to 9% over the next six months.

The reality is many firms are still figuring out how to fit AI into their processes.

‘With AI handling tax prep, you'll have more capacity than ever. You can focus on client conversations that actually build relationships and trust. With A.I., you'll have the right tools to engage in advisory services where you'll be able to provide a bigger value to clients, whether it's in tax savings or wealth preservation.’

Kevin Lee avatar
Kevin Lee
CPA, Filed

Experts recommend getting your team involved early by asking senior staff to vet AI-prepped returns and gather feedback on what works and what doesn’t. When they see how the tech can catch errors, organize client papers, and speed up reviews, resistance tends to fade fast.

The AI adoption survival guide

Staff concern
‘Am I about to become an AI engineer instead of a tax pro?’
Reality check

A CPA flags an issue on a return processed by an AI tool. The AI doesn’t just move on; it remembers. It refines its accuracy, integrating real-world tax data and CPA feedback so that the same issue doesn’t show up again next year.

Staff concern
‘Will I get blamed if the AI gets it wrong?’
Reality check

AI doesn't change the chain of accountability. Just like you’d never file a return prepped by a junior associate without reviewing it, the same goes for AI. The difference? AI gets you to that review stage faster (with fewer manual errors to clean up).

Staff concern
‘I’ve spent years perfecting my method, do I really need to change it?’
Reality check

You’re not changing your process; you’re optimizing it. The best AI-powered tax prep tools integrate directly with your existing software, so your workflow doesn’t have to go out the window.

Staff concern
‘Will AI make my work feel less personal or client-focused?’
Reality check

Actually, it’s the opposite. With data entry and document wrangling handled automatically, you’ll have more time to build relationships and offer strategic advice that keeps clients loyal.

‘It's really this idea of working with AI instead of being replaced by it. Tools like Filed still have internal CPA experts handling the advanced cases. It's a partnership.’

Sacharias Sjöqvist avatar
Sacharias Sjöqvist
AI & ML Software Engineer, Filed

Communicating with clients

Even if your team’s on board, your clients may still need reassurance that they’re getting the same level of service (if not better).

The truth is, most of them don’t care whether AI is involved; they just want to know their return is accurate, secure, and their CPA is in control.

The best client communication highlights how AI improves the tax filing process. It’s an extra layer of accuracy and protection, and the human element hasn’t gone anywhere: your clients are still getting the same level of attention and expertise–and now they benefit from smarter tools that make sure nothing slips through the cracks.

Pro tip: Smart firms are implementing ethical disclosure policies that openly explain how they use AI and its role in processing client data.

At the end of the day, clients aren’t impressed by how sophisticated your tech stack is. They’re impressed by how much tax you save them.

Two tax professionals having a conversation over coffee.

‘I don't think clients would have an issue with CPAs using AI. In fact, I'd emphasize it, saying 'We use AI so our workflows are more efficient.' I honestly think it would be more of an issue if you told clients you offshore the work–most wouldn't be comfortable knowing their financial information is being processed overseas with limited security oversight.’

Mohson Choudhry avatar
Mohson Choudhry
CPA, Ayub Accounting
Force #5

The reinvention of the tax pro

Man in a suit holding red file folders.
<90%
of IT leaders turn to offshore solutions, especially in peak season.

And that shift brings both opportunity and risk. Firms that embrace it are finding new ways to grow, while those who resist are stuck competing on price instead of value.

[ CPAs finally have the space to focus on higher-value, client-centric services. ]

35%
of accountants say they shift toward advisory

The industry is already moving this way. Advisory services are becoming the norm. In fact, 35% of accountants say the shift toward advisory has been the most significant industry change in the past five years. That’s more than double the impact of AI (13%) and cloud technologies (15%).

Tax prep alone isn’t enough to run a successful firm anymore, especially with recruitment and economic pressures mounting.

CPAs using AI aren’t just speeding up data entry; they’re changing how they do business. Instead of clocking hundreds of tax prep hours and retraining junior seasonal staff every year they’re focusing on advisory services: tax efficiency planning, wealth strategies, and even fractional CFO work.

But dialing up your advisory services isn’t as simple as flipping a switch.

This transition tends to be smoothest for firms with strong client relationships, deep industry expertise, and a willingness to invest in their team’s development. Those who build on their strengths rather than trying to change everything at once are seeing the best results.

‘CPAs learn a lot about their clients through doing their tax over many years. There’s a lot of valuable advice outside of prepping returns.’

Mohson Choudhry avatar
Mohson Choudhry
CPA, Ayub Accounting

But remember, AI isn’t a magic bullet.

While AI tools like Filed can automate the grind of data entry, they aren’t a universal fix. The most successful firms know that AI is a tool, not a takeover. They’re building smarter workflows and guardrails while upskilling their team to balance automation and expertise.

These CPAs are developing a strategic approach that:

1) Uses AI for routine tasks, slashing payroll costs without the annual scramble to find, hire, and train seasonal staff.

2) Leverages AI’s memory of firm workflows and client context, so they’re not starting from scratch every year (even if their team is).

3) Keeps human expertise for complex scenarios.

4) Continuously refines technological capabilities, making each tax season more predictable than the last.

5) Invests in training to maximize AI collaboration, creating a workplace where burnout becomes the exception, not the rule.

AI oversight
Learning how to keep the tech in check, understanding how AI processes tax data, and knowing when human judgment is essential.
Data-driven insights
Getting comfortable translating financial data into actionable strategies like entity structuring and risk assessments.
Client communication
Spotting when clients need help understanding tax positions and legislation changes.

‘The biggest theme in successful firms right now? CAS—Client Advisory Services. If you’re running a bakery or a garage, you don’t have time to dig into your financials. That’s what CPAs are for—but only if tax prep and other basics are handled smoothly.’

Leroy Kerry avatar
Leroy Kerry
CEO, Filed

The skepticism spectrum

Take the quiz to find out where you stand

Headless man in suit with a gear icon as head.
Traditionalist
‘If It ain't broke...’

You've built a successful practice the old-fashioned way: with grit, expertise, and possibly actual paper files. Tech comes and goes, but tax laws are forever.

What CPAs like you are thinking about:

  • What specific aspects of my practice are truly working well vs areas where I'm feeling pinch points.
  • Whether clients are starting to inquire about tech-enabled services.
  • How to evaluate claims about security and compliance in ways that align with my standards.
Headless man in trench coat with magnifying glass and eye icon as head.
Cautious Observer
‘We'll wait and see’

You've seen enough flash-in-the-pan tax tech to know better than to jump on every bandwagon. But you're keeping an eye on this one.

What CPAs like you are thinking about:

  • Which admin tasks consistently steal time from higher-value work.
  • How to identify which AI claims are marketing hype vs substantiated benefits.
  • What question marks need answering before a more serious evaluation of AI makes sense.
Headless man in tuxedo giving thumbs up with spinning top icon as head.
Experimenter
‘Let's test the waters’

You're practical but progressive. New tools get a fair trial, but they'd better prove themselves quickly.

What CPAs like you are thinking about:

  • Which return types present the lowest risk for initial testing.
  • What success metrics would justify broader implementation.
  • How to maintain quality control throughout experimentation.
Headless man in checkered blazer with puzzle piece icon as head.
Integrator
‘AI is a tool, not a threat’

You see AI as another extension of your professional toolkit, like moving from paper to spreadsheets, but with more upside.

What CPAs like you are thinking about:

  • How to identify the highest-impact integration points in your workflow.
  • What client communication strategies work best when introducing AI-assisted services.
  • Which team members show the most aptitude for bridging tech and tax.
Headless man in suit with lightbulb icon as head.
AI-powered firm
‘Work smarter, not harder’

You're building your firm's future around AI. While others debate, you're busy reinventing tax practice.

What CPAs like you are thinking about:

  • Refining your AI implementation based on performance data.
  • Developing distinctive service offerings that leverage your tech advantage.
  • Balancing automation with the human expertise clients still value highly.

AI-skeptic 
to AI-ready

Man working on a laptop.

Whether you're a cautious observer dipping your toe in the water or an integrator ready to scale, this roadmap will help you test, implement, and optimize AI without losing control over decision-making.

‘Don’t just hand off tax prep to AI and assume it’s perfect. Keep doing your due diligence—make sure AI gets all the right documents upfront. You’re still responsible for accuracy, and that’s where your expertise makes the difference.’

Kevin Lee avatar
Kevin Lee
CPA, Filed

A CPAs step-by-step AI adoption roadmap

Step 1: Start small, win big
Finding low-risk, high-impact use cases.

Why: The smartest AI implementations don't try to boil the ocean. They target specific pain points where automation delivers immediate value.

Analyze last season's workflow
  • Which tasks consumed the most staff hours?
  • Where did quality issues or bottlenecks occur?
  • Which processes did you dread?
  • Or which generated the most staff complaints?
Identify repetitive, time-consuming tasks
  • Target the grunt work first (document sorting, data entry from standard W-2, 1099 forms).
  • Choose high-volume, standardized returns (1040, 1065, 1120, 1120S).
  • Leave your most complex, judgment-heavy returns for humans.
Create a safety net
  • Select a test group of 10-15 returns.
  • Run parallel processes until you build confidence.
  • Compare AI-prepped returns with traditionally prepped ones.
Get your team on board
  • Involve your most skeptical staff members (they ask the best questions).
  • Frame AI as eliminating busywork, not expertise.
  • Provide training that builds confidence in working alongside AI.
Step 2: Keep control
Maintaining professional standards while automating.

Why: Your professional judgment is what clients pay for. 
AI should enhance it, not replace it.

Design a smart review system
  • Start with 100% CPA verification of AI outputs.
  • Create clear criteria for which returns require deeper scrutiny.
  • Gradually shift to risk-based review as confidence grows.
Define professional judgment boundaries
  • Map exactly where AI hands off to human expertise.
  • Document instances where professionals override AI suggestions.
  • Develop a feedback loop to improve the human-AI workflow.
Prepare for client questions
  • Decide how (or if) to disclose AI usage to clients.
  • Prepare client talking points that emphasize quality benefits.
  • Train staff on addressing potential client concerns.
Stay compliant
  • Document all AI-assisted processes for potential IRS review.
  • Establish clear responsibility chains for accuracy.
  • Maintain version control of AI systems throughout tax season.
Step 3: Scale without sweating
Reinvest your time where it matters.

Why: The real ROI of AI isn't in cost reduction alone. 
It's in transforming your practice to deliver higher-value services.

Measure what matters
  • Track time saved, error rates, and client satisfaction.
  • Calculate the actual dollar value of reclaimed hours.
  • Identify which processes improved most (and which still need work).
Deploy your newfound capacity strategically
  • Expand into higher-margin advisory services.
  • Take on more clients without hiring seasonal staff.
  • Develop expertise that AI can't match (and clients will pay for).
Evolve your pricing model
  • Consider value-based pricing for
AI-enhanced services.
  • Develop year-round service packages beyond tax season.
  • Bundle compliance work with the strategic advice clients actually values.

The AI-powered tax practice: what’s next?

Three people having a business meeting.

The tax profession stands at a crossroads of necessity and opportunity.

This report has dissected the five forces reshaping tax in 2025, from data security to evolving economics and the reinvention of the CPA role itself, but the most profound insight emerged from between the lines: AI's true potential isn't in automation, but transformation.

Instead of just eliminating the busy work that drains your time and energy, AI creates a ripple effect throughout your entire practice.

Preparers evolve into reviewers, managers power through organized reviews instead of hunting for errors, and partners spend less time supervising and more time advising—all while clients get faster turnarounds and deeper strategic insights. You reclaim hours of your time and gain complete control over when and how to grow your profits.

[ AI isn’t a magic wand that will turn your firm into a futuristic powerhouse overnight. ]

The firms gaining an edge aren’t waiting for a mythical ‘perfect moment’ to jump in, but they aren’t going all in on automation just because it’s trendy, either.They’re trialing it, testing it, and figuring out how to make it work in a way that sticks.

They’re drawing their own line in the sand: they know when to lean into AI and when to lean on their expertise.

But this isn’t the end of the story; far from it.

The next wave of innovation is making AI even more adaptable and practical.

Imagine AI agents that don’t just automate tasks but spot patterns in financial data, research complex tax scenarios, and pull together insights that make your advisory work sharper and faster.

Or on-premises solutions that keep the power of AI right within your office. No sending sensitive data to third-party servers (even US ones) or trade-offs between innovation and control.

[ As AI tools get smarter, the tax pro’s role will keep evolving, too. ]

The tax pro of tomorrow isn't just a tax code interpreter. They'll be deploying the most powerful tools available to keep pace with the demands of modern practice and build relationships that stand the test of time.

‘The vision has always been to give CPAs back time, whether it's for the seasoned pro David, who wants more time with his grandkids, or the partner who wants to do more work with clients than paperwork. Our AI is about turning low-value tasks into opportunities to improve top-line growth by doing more high-value services.’

Leroy Kerry avatar
Leroy Kerry
CEO, Filed