The federal noncompete ban is over as a rulemaking. The FTC agreed to vacate the Non-Compete Clause Rule and drop its appeals in September 2025, and on February 12, 2026 published a final rule formally removing it from the Code of Federal Regulations. A district court in Ryan LLC v. FTC had already blocked the rule nationwide days before it was to take effect.

What replaced it is not deregulation. The FTC signalled after a January 2026 workshop that it would abandon the categorical ban in favour of case-by-case enforcement under Section 5 of the FTC Act, with attention on agreements covering lower-level employees and on clauses that are exceptionally broad. Meanwhile the states kept legislating.

So the practical answer to “is my noncompete enforceable?” is what it was before the rulemaking started: it depends on the state, and increasingly on the employee’s income.

Where the lines are drawn

Four states will not enforce employee noncompetes at all: California, Oklahoma, North Dakota and Minnesota. Washington becomes a near-total ban on June 30, 2027.

Beyond the outright bans, the significant trend is income thresholds — Colorado and Washington among others restrict enforceability below a stated compensation level, indexed annually. The practical effect is that noncompetes have become an executive-and-key-employee instrument rather than a workforce-wide one, and a clause applied to everyone is now the specific fact pattern that draws both state scrutiny and the FTC’s stated Section 5 interest.

For a mid-market employer with staff in several states, this means one template cannot be correct everywhere. A blanket clause is unenforceable in some of your states, enforceable in others, and void-or-blue-pencilled in the rest depending on how that state’s courts handle overbreadth.

Consultants are a different question entirely

A noncompete binding an independent contractor is a different legal animal from one binding an employee, and the differences run in both directions.

Consideration is the first issue. In many states continued employment is adequate consideration for an employee noncompete; for a contractor, the analysis is contractual and the absence of separate consideration is a real vulnerability.

Scope is the second. A consultant whose business is serving multiple clients in one industry cannot be barred from that industry without effectively ending their livelihood, which is the fastest route to a court declining to enforce. Clauses that survive tend to be narrow: this client’s specific customers, for a defined period, in a defined role.

And the classification risk is the one that stings. A noncompete restricting where and for whom someone may work is precisely the kind of control that supports reclassifying them as an employee. Firms that use aggressive noncompetes with contractors sometimes win the noncompete argument and lose a much more expensive classification argument.

What actually protects the business

Most of what employers want from a noncompete is available through instruments that are enforceable nearly everywhere and generate far less litigation.

Confidentiality and trade-secret protection. Enforceable in every state, with statutory backing. It protects the actual asset — the information — rather than restricting a person’s employment.

Customer non-solicitation. Narrower, more defensible, and usually what the concern really was. Limit it to customers the person actually served, and to a period that reflects real relationship decay, not the longest term you think you can get.

Employee non-solicitation. Generally enforceable if reasonable, though a few states now scrutinise these too.

Notice periods and garden leave. Underused in the US. A paid notice period achieves the same protective effect as a short noncompete, with the crucial difference that you are paying for the restriction, which courts view very differently.

Sensible equity and deferred-compensation terms. Forfeiture on competition is treated more permissively than an outright employment restriction in many states — though not all, and not in California.

Choice-of-law will not rescue a bad clause

The common workaround is to write the agreement under the law of a state that enforces noncompetes and hope that governs. It mostly does not work where it matters. States with strong public policy against these clauses — California most aggressively — will apply their own law to a resident employee regardless of what the contract says, and several have added statutory provisions that void out-of-state choice-of-law and venue clauses in employment agreements outright.

The practical reading: choice-of-law is useful for the genuinely ambiguous middle, and useless against a state that has decided the question. If a person lives and works in a ban state, the clause is unenforceable there no matter which state’s law the template names, and attempting to force the issue in another venue is expensive litigation with a predictable ending.

The audit worth doing this quarter

Pull every template containing a restrictive covenant: offer letters, employment agreements, contractor agreements, equity documents, settlement and severance agreements. For each, record which state’s law governs, which states the covered people actually work in, and what the income threshold is in those states.

Then delete the ones you cannot enforce. An unenforceable noncompete is not free — it deters people who do not know it is void, misleads managers into believing they have protection they do not, and in several states now carries its own penalty for including it in the first place.

Reserve real noncompetes for the small set of people whose departure genuinely threatens the business, pay separate consideration for them, keep the scope narrow enough that a judge reading it thinks it sounds reasonable, and run the state-by-state check before the offer letter goes out rather than after the resignation.