Independent Contractor Agreement Red Flags: What to Check Before You Sign

1. Classification Language Versus the Actual Relationship

Every independent contractor agreement states, in some form, that you are an independent contractor and not an employee. That sentence matters far less than whether the rest of the agreement, and the actual working relationship it describes, is consistent with it. Worker classification is governed by state and federal tests that look at the real facts, not the label a contract chooses, and the tests themselves vary by state and by which law is being applied, so read the working conditions the agreement describes rather than trusting the heading alone.

Watch for language that puts the client in control of your schedule, requires you to use the client’s equipment and workspace, bars you from working for anyone else during the engagement, or describes ongoing, indefinite work rather than a defined project. None of these, alone, reclassifies you, but a cluster of them in the same agreement is worth a second look, both for your own protection and because a misclassification dispute can unwind the engagement for both sides.

What a reasonable version looks like: a scope, schedule, and level of control consistent with genuine independence, and a clear sense of how classification is treated in the state where the work happens, since the applicable test is not the same everywhere.

2. Scope of Work and Change Orders

The scope of work is supposed to be the boundary of what you owe for the agreed rate. A vague scope, phrased as ongoing support, reasonable additional tasks, or work as the client directs, leaves that boundary undefined, and undefined boundaries tend to expand in the client’s favor over the life of an engagement. A specific scope, naming deliverables, milestones, and what falls outside them, is what actually protects the rate you negotiated.

Just as important is what happens when the scope needs to change, since most real engagements do change. A change order process, a written description of new work and its added cost or timeline, agreed before you start it, keeps scope creep from becoming unpaid work. An agreement silent on this point tends to resolve every ambiguous request in the client’s favor by default, since you have nothing in writing to point to.

What a reasonable version looks like: a specific, itemized scope of work, plus a written change order process for anything added to it, with the added cost or time agreed before the new work starts.

3. Payment Terms and Late Fees

Confirm the rate, the billing cycle, and the actual payment window, commonly net fifteen, net thirty, or net forty-five days from invoice. A longer payment window is not automatically unreasonable, but it changes your cash flow planning, particularly if the engagement is your primary income. Also confirm what triggers an invoice: a calendar date, a milestone, or the client’s own approval of the work, since a client-approval trigger without a deadline can quietly stall payment indefinitely.

Check whether the agreement includes a late fee or interest provision for a payment that misses its window, and whether it gives you the right to pause work if payment is significantly overdue. An agreement that specifies exactly how the client will pay late but says nothing about what happens if the client simply does not pay is more common than it should be.

What a reasonable version looks like: a defined payment window from invoice or milestone, a specific invoicing trigger, and a stated late fee or a right to pause work if payment is materially overdue.

4. Kill Fees and Termination for Convenience

Most independent contractor agreements let the client end the engagement for convenience, meaning for any reason or no reason, usually on written notice of some number of days. That is ordinary and expected. What varies widely, and what decides whether an early termination costs you real income, is whether the agreement includes a kill fee, payment for work in progress or for a stated percentage of the remaining engagement, if the client cancels after you have already turned down other work to take this one.

Read the notice period as closely as the fee itself. A same-day termination clause with no kill fee means you can lose an entire project’s worth of planned income with no warning and no compensation for the gap.

What a reasonable version looks like: a stated notice period for termination without cause, payment for all work completed and in progress through the termination date, and a kill fee or minimum notice period substantial enough to protect against losing other work you turned down for this engagement.

5. IP Assignment Versus License, and Portfolio Rights

Read exactly what happens to the work product you create. An assignment clause transfers full ownership of the finished work to the client, which is standard for most client-facing deliverables and is not, by itself, something to push back on. A license clause instead lets the client use what you created without you giving up ownership, which is more common for tools, templates, or methods you built independently and are applying to this engagement rather than creating from scratch for it. The difference matters most for anything you intend to reuse on future projects.

Separately, check whether the agreement lets you show the finished work in a portfolio or case study, and whether that permission requires the client’s prior approval, a delay after project completion, or is silent on the point entirely. A silent agreement defaults, in most jurisdictions, to the client’s ownership controlling the question, which can mean you legally cannot show your own work to a future client without asking first.

What a reasonable version looks like: a clear statement of assignment versus license for each category of work product, an explicit carve-out for your own pre-existing tools and methods, and a stated right to use the finished work in a portfolio, even if that right requires notice or a short delay.

6. Moral Rights and Work-Made-for-Hire Wording

Some agreements state that your work is a work made for hire, a specific legal category that, for certain kinds of work, makes the client the legal author from the moment of creation rather than transferring ownership after the fact. Whether that category actually applies depends on the type of work and the details of the engagement, not just the label the agreement uses, so a work-made-for-hire clause is often paired with a backup assignment clause in case the category does not apply. Read for both, since an agreement relying on the label alone without a backup assignment can leave ownership genuinely unclear if the category is later challenged.

A separate, narrower provision sometimes asks you to waive moral rights, your personal right to be credited for the work and to object to a distorted or modified version of it. This is standard in many creative and technical fields and is not, on its own, unusual, but it is worth reading in full rather than skimming past because it sits near the ownership language.

What a reasonable version looks like: a work-made-for-hire clause paired with a backup assignment clause covering the same work, and moral rights language scoped to what the engagement genuinely requires rather than an unlimited waiver.

7. Indemnification and Its Caps

An indemnification clause says who pays if a third party brings a claim connected to the work, a client sued because your deliverable infringed someone else’s copyright, for example. A one-sided clause that makes you indemnify the client for essentially anything related to the engagement, with no dollar limit, exposes you to liability far larger than what the engagement itself is paying you. A mutual clause, where each party indemnifies the other only for claims actually caused by that party’s own conduct, is the more balanced version.

Look specifically for a cap tying your indemnification obligation to the fees actually paid under the agreement, or to some other defined ceiling, rather than leaving it open-ended. An open-ended indemnification obligation on a modest engagement is a mismatch between the risk you are accepting and the amount you are being paid to accept it.

What a reasonable version looks like: mutual indemnification, scoped to each party’s own conduct and to third-party claims genuinely connected to the work, with your obligation capped at the fees paid under the agreement.

8. Limitation of Liability

A limitation of liability clause caps how much either party can be forced to pay in damages if something goes wrong, separate from indemnification, which covers third-party claims. Most reasonable agreements cap your liability at the fees paid under the agreement, and exclude indirect, consequential, and speculative damages, such as a client’s claimed lost profits from a delayed launch. An agreement with no liability cap at all leaves you exposed to a damages claim that could exceed the entire value of the engagement many times over.

What a reasonable version looks like: a mutual liability cap, tied to fees paid, that excludes indirect and consequential damages, applied to both parties rather than only to you.

9. Insurance Requirements

Many client agreements, particularly with larger companies, require you to carry a specific type and amount of insurance for the length of the engagement, commonly general liability, professional liability or errors and omissions coverage, and sometimes cyber liability if you handle the client’s data. These requirements are ordinary business practice, not a red flag on their own, but the required coverage amount should be proportionate to the size and risk of the engagement rather than a boilerplate figure copied from a much larger contract.

Confirm the cost of the required coverage before you sign, since it is a real expense that reduces your effective rate on the engagement, and factor it into whether the rate you negotiated still works once that cost is accounted for.

What a reasonable version looks like: insurance requirements proportionate to the size and risk of the engagement, stated clearly enough that you can price the coverage before signing rather than discovering the cost afterward.

10. Non-Compete and Non-Solicit Riders

An independent contractor agreement is supposed to govern one engagement. Some agreements also include a non-compete clause, restricting you from working with the client’s competitors, or a non-solicit clause, restricting you from pursuing the client’s customers or employees, sometimes for a year or more after the engagement ends. As an independent contractor who typically works with multiple clients, a broad non-compete can meaningfully limit your ability to earn a living once this one engagement is over, which is a materially different situation than an employee’s non-compete.

A narrow non-solicit tied to the specific clients or employees you actually worked with during the engagement is common and generally reasonable. A broad non-compete covering an entire industry or client base you never touched deserves real pushback, since it can outlast the value of the engagement itself.

What a reasonable version looks like: no broad non-compete, and a non-solicit, if any, narrowed to the specific clients, prospects, or employees you actually worked with, for a defined and reasonable period.

11. Confidentiality and How Long It Survives

Confidentiality provisions in a contractor agreement work the same way they do in a standalone NDA: they should define what counts as confidential, exclude information that was already public or that you already knew, and state how long the duty lasts after the engagement ends. A duty that survives indefinitely, with no end date, asks you to guess forever whether something you learned years ago on a short project is still covered.

Check as well whether the confidentiality obligation is mutual, since you may share your own methods, rates, or business information with the client over the course of the engagement, and a one-way clause protects only the client’s side of that exchange.

What a reasonable version looks like: a defined survival period, typically one to three years past the engagement’s end for ordinary business information, standard exclusions for public or independently known information, and mutual obligations where both sides actually exchange sensitive information.

12. Exclusivity

An exclusivity clause restricts you from taking on other clients, either in general or within a specific industry or category, during the engagement. This is a meaningfully bigger commitment than it may first appear for a contractor whose income typically depends on multiple concurrent clients, and it deserves compensation proportionate to the income you are giving up the ability to earn elsewhere.

If exclusivity appears in a draft and was not part of the rate discussion, treat it as a term to negotiate rather than accept by default, since agreeing to it quietly changes the actual economics of the engagement.

What a reasonable version looks like: no exclusivity requirement unless it was priced into the rate, and if it is required, a scope narrow enough, by industry or by specific competing clients, that it does not block your other income.

13. Acceptance and Revision Limits

Read how the agreement defines a finished, accepted deliverable. A workable clause states a review period, commonly five to ten business days, after which the deliverable is deemed accepted if the client raises no objection, and a defined number of revision rounds included in the rate before additional revisions are billed separately. Without an acceptance clock or a revision limit, a client can withhold approval indefinitely and request unlimited revisions at no additional cost, which quietly turns a fixed-price project into open-ended, unpaid work.

What a reasonable version looks like: a stated review period with deemed acceptance if the client does not respond, and a defined number of included revision rounds, with any revisions beyond that billed at your standard rate.

14. Expenses

If the engagement involves any cost beyond your own time, travel, software, subcontracted work, or materials, confirm whether those expenses are reimbursed, and if so, whether reimbursement requires pre-approval, a receipt, or a specific submission process and deadline. An agreement silent on expenses is usually read as meaning you absorb them, which is a real reduction in what the engagement actually pays you once those costs are counted.

What a reasonable version looks like: a stated expense reimbursement policy for any cost beyond your own time, with a clear pre-approval and submission process rather than silence on the point.

15. Governing Law, Venue, and Who Pays for Disputes

The governing law clause decides which state’s law a court or arbitrator applies to the agreement, and the venue clause decides where a dispute is heard. A client based far from you may propose its own home state and city for both, which raises the practical cost of ever enforcing your own rights under the agreement, even when you are clearly right on the facts.

Also check who pays attorney fees and dispute costs if a disagreement escalates. A one-sided fee-shifting clause, where you pay the client’s legal costs if the client wins but the client never pays yours, tilts the risk of any dispute heavily toward whichever side wrote the agreement, which is nearly always the client.

What a reasonable version looks like: a governing law and venue with a real connection to the engagement, such as your own state or a neutral one, and mutual fee shifting or none at all rather than a one-sided version.