Working mostly for one client? The rule changed. The risk did not.

A client called me last fall, panicked. She had been working with a single software company for fourteen months, almost full time, billing through her LLC. The IRS had sent her client a letter. Could the company have been treating her as a contractor when she should have been an employee?

The answer, which I have given some version of to a lot of clients in the last few years, was: maybe, and the test that determines it is more about the working pattern than the contract.

This is the part of independent contractor classification that gets the least attention and creates the most actual risk. The forms can be perfect. The contract can be tight. The invoices can be regular. None of that protects you if the day to day working relationship looks like employment.

This is general information, not legal advice. Specifics depend on your state, your industry, and your facts. Talk to an actual lawyer before you act.

The current rule, briefly

In early 2024, the Department of Labor under the Biden administration issued a final rule on independent contractor classification under the Fair Labor Standards Act. It used a six factor "economic realities" test, with no factor weighted more heavily than the others. It was widely seen as making it harder to classify workers as contractors.

In February 2026, the current Department of Labor issued a Notice of Proposed Rulemaking that would replace the 2024 rule with something closer to the 2021 rule. It still uses an economic realities test, but with two factors weighted as "core": the nature and degree of control over the work, and the worker's opportunity for profit or loss based on initiative and investment.

The comment period closed in late April 2026. As of May, the 2024 rule is still the operative rule for federal Fair Labor Standards Act enforcement. The new rule will not take effect until a final version is published, which is probably late 2026 at the earliest.

The political back-and-forth is real and matters at the margin. But the broader point I want to make is that for most freelancers, the change in the federal rule is not the thing that determines their risk. The IRS uses a different test. Most states use their own tests, some of which (California's AB5, for example) are much stricter than any federal rule. And the practical patterns that trigger trouble look the same under almost any test.

What actually triggers a classification problem

In my experience helping freelancers navigate audits and disputes, the formal legal test almost never matters as much as the day to day pattern. The pattern that creates problems looks like this:

1. The freelancer works for one client, or one client dominant relationship that is 70% or more of their income.
2. The work is ongoing rather than project based. The freelancer is integrated into the client's regular operations.
3. The client provides the tools, the systems, the email account, the laptop, or the workspace.
4. The freelancer's schedule is set by the client, often with required meetings or "core hours."
5. The freelancer is given direction on how to do the work, not just what to deliver.
6. The relationship has lasted more than a year, and there is no clear project endpoint.

If two or three of these are true, you have a relationship that any auditor would scrutinize regardless of which rule applies. If five or six are true, you have an employment relationship that is being legally treated as a contractor relationship, and the only question is whether anyone notices.

Notice what is not on this list. Whether you have an LLC. Whether the contract calls you an independent contractor. Whether you send 1099-eligible invoices. Whether you have a written agreement at all. All of these matter at the margin. None of them override the working pattern.

The legal phrase for this is the substance over form doctrine. What you actually do controls how the relationship is classified, not what you call it.

Why this matters for the freelancer

People assume the classification risk is on the client. It is mostly on the client, but it is not only on the client.

If a freelancer is reclassified as an employee retroactively, the client owes:

1. Back payroll taxes (the employer half of FICA, plus FUTA, plus state unemployment).
2. Potentially back wages if the worker was making less than minimum wage when divided by hours worked.
3. Benefits, if the company has a benefits plan that the worker should have been enrolled in.
4. Penalties and interest.

The freelancer, in the same situation:

1. Owes income tax adjustments depending on how their previous filings interact with the recharacterized income (often complicated, sometimes net neutral, sometimes painful).
2. Loses the self employed health insurance deduction, the home office deduction, and other Schedule C expenses for the relevant years.
3. May have to renegotiate their own retirement contributions if a solo 401(k) was funded based on what is now W2 income.
4. Almost certainly loses the client, because the client now has a reason to never work with them again.

The freelancer rarely owes massive back taxes, but they often lose more in expense deductions than they saved by being a contractor in the first place. And the client relationship is functionally over.

The contract changes that actually help

The contract is the easy part. A few specific clauses make a meaningful difference.

1. Scope, scope, scope. Define the work as a project or series of projects with clear endpoints. "Ongoing support" is the phrase that gets you into trouble. "Three month engagement to deliver X, Y, and Z" is the phrase that protects you.
2. Right to subcontract. A contractor can hire someone else to do the work. An employee cannot. The contract should explicitly grant you the right to subcontract or delegate parts of the work.
3. Right to other clients. The contract should not contain exclusivity. If a client insists, push back hard or recognize you are pricing the engagement as if it were employment.
4. Your own tools. Your work product is delivered using your own laptop, software, and infrastructure. If the client wants you to use their tools, you negotiate a tooling fee or factor it into the rate.
5. Set your own schedule. No required hours. No required attendance at recurring internal meetings. If meetings are needed, they are mutually scheduled, like a vendor would be.
6. Fixed fee or milestone billing. Avoid pure hourly time tracking if you can. A retainer for a defined deliverable scope is fine. Per-hour billing without a scope cap starts to look like employment.

None of these are bulletproof, but each one makes the classification argument easier if it ever comes up. They also make the day-to-day relationship cleaner, which is usually good for both sides.

The pattern changes that matter more

The contract is the floor. The behavior is what actually decides the case.

If you are working with one dominant client, here are the patterns that protect you:

1. Have other clients. Even small ones. The freelancer who has one client at 95% of revenue and the freelancer who has one client at 70% of revenue plus three smaller engagements look very different to an auditor. Real other clients, paying real invoices.
2. Set your own hours. Pick up your kids at 3 PM. Take Fridays off. Work from a coffee shop sometimes. The point is not to be unreliable. The point is to demonstrate that the client does not control your schedule.
3. Decline some requests. If the client asks for something outside scope, treat it as a change order with a fee attached, not as a quiet "yes" to keep the relationship friendly. Employees say yes. Contractors negotiate.
4. Invoice formally. Don't just submit hours. Invoice with your business name, your terms, and your details. Each invoice is a small assertion that you are running a business.
5. Have a public presence as a business. A website. A LinkedIn profile that says you are a freelancer. Recent posts or work shown elsewhere. The auditor scanning your situation should see a person running a business, not a person playing the role of an employee.

These are unsexy and slow, and almost all of them are good business hygiene regardless of classification risk. They also do most of the actual work of keeping you on the right side of the line.

The long term single client question

The harder question, which the rule changes do not really address, is whether you should be in a long term single client engagement in the first place.

Sometimes yes. Some clients are large enough and the work is meaty enough that 70 or 80% concentration is hard to avoid, especially if you have specialized into something narrow. The compensation is often very good. The work is interesting. The relationship is real.

But the concentration is itself a business risk separate from the classification question. The client can end the relationship. Your industry can shift. The point person at the client can change jobs. The classification risk is just one of the ways a single client engagement can blow up, and not even the most common one.

The advice I give freelancers in this position is the same advice I give any business owner with concentrated customer revenue. Diversify when you can. Negotiate carefully when you cannot. Treat the engagement as a contract that may end at any time, and structure your finances and your career around that possibility.

The classification rule is mostly a question of paperwork. The single client risk is a question of running a real business. The first is a problem you can solve with a lawyer. The second is one you have to solve yourself.