How New York Courts Treat Restrictive Covenants After the FTC Rule

The FTC's 2024 non-compete rule made national headlines. The headline kept producing headlines after a Texas federal court blocked it. What got less attention is that, while the rule itself was in limbo, New York trial courts were quietly recalibrating how they handle restrictive covenants — almost always in the employee's favor. Eighteen months in, the trend is clear enough to draft around.

If you're an employer relying on non-competes, non-solicits, or confidentiality clauses to keep talent and customers, the playbook has changed. Here's what to know going into 2026.

The trend lines

In the eighteen months since the FTC announced its rule, New York trial courts have denied preliminary injunctions in restrictive-covenant cases at the highest rate in over a decade. The denial rate, by my unscientific count of reported decisions, runs around 60-65% — up from a historical baseline closer to 40%. Some of that is selection effect: cases that would have been easy injunction grants pre-2024 are now being settled pre-suit because employers know the bar has risen.

The shift is not statutory. The FTC rule never took effect in New York. New York's state-level non-compete ban was vetoed. The shift is judicial — and it's real.

What the courts are actually saying

Three themes run through the post-2024 New York opinions denying preliminary injunctions in restrictive-covenant cases:

Theme 1: Irreparable harm is harder to prove. Pre-2024, employers could often establish irreparable harm by showing that the former employee had access to customer relationships or confidential information. Post-2024, courts demand specificity. Which customers? What confidential information? How exactly will it be used by the new employer? Without granular evidence, the irreparable-harm prong fails.

Theme 2: Legitimate business interest is narrower. The judicial concept of a "legitimate business interest" has tightened. General competitive disadvantage doesn't count. Loss of investment in training doesn't count without specifics. The interest must be in specific customer relationships the employee personally serviced, specific trade secrets the employee accessed, or specific irreplaceable contributions the employee made.

Theme 3: Reasonableness is judged against modern employment markets. Courts are increasingly willing to point out that the employer's business will not collapse if one mid-level employee takes a job at a competitor. The pre-pandemic intuition that any departure to a competitor inflicts "irreparable harm" has been quietly retired.

The "blue-pencil" problem

Some New York judges will narrow an overbroad clause to make it enforceable. Others won't. Increasingly the trend in the Commercial Division is to refuse to rewrite. The reasoning: if employers know courts will save them by narrowing overbroad clauses, employers have no incentive to draft reasonably in the first place.

This matters for drafting. The old strategy of "ask for everything, take what the court will give" is much riskier in 2026. A clause that's 30% too broad gets struck entirely. The smart strategy is to draft narrowly and defensibly from the start — and to use multiple, severable clauses (non-compete, customer non-solicit, employee non-solicit, confidentiality) so that even if one is struck, the others survive.

Customer non-solicits are the new battleground

Non-competes get the headlines. But the harder fight in 2026 is over customer non-solicits, especially where the departing employee brought the customer relationships into the company in the first place.

The classic fact pattern: senior salesperson joins a B2B company, brings their book of business, services those customers for five years, then leaves. The non-solicit prevents them from approaching their own customers for 18 months. Courts have grown skeptical. The recent trend is to read non-solicits to cover only customers the employee acquired or substantially developed through the employer's resources, training, and goodwill — not customers the employee brought in.

Drafting fix: distinguish in the contract between "employee-originated" customers (carved out of the non-solicit) and "company-originated" customers (covered by the non-solicit). It's a harder negotiation at hiring, but it produces a clause that actually holds up.

Employee non-solicits: the underrated weapon

While non-competes are weakening, employee non-solicits have held up surprisingly well. The legitimate business interest in workforce stability is easier to prove than the legitimate business interest in customer retention. Most New York courts will enforce a reasonable employee non-solicit (12-18 months, limited to employees the departing employee actually worked with) without much friction.

For employers, that means: if you're going to fight one restrictive-covenant battle in 2026, fight the employee non-solicit. It's more likely to be enforced and the damages are easier to prove.

The confidentiality clause is the workhorse

Confidentiality clauses have always been more enforceable than non-competes, and that gap has widened. A well-drafted confidentiality clause — specific about what counts as confidential, what doesn't (publicly available information, information the employee knew before joining), and the duration of the obligation — is the most reliable post-employment protection an employer has.

Two drafting points often missed: (1) carve out information that becomes public through no fault of the employee, and (2) include a return-of-materials clause requiring the employee to certify in writing, on departure, that they have returned or destroyed all company-confidential materials. The certification is procedurally simple and gives the employer a clear cause of action if it turns out the employee kept files.

The federal-court vs. state-court split

Some New York federal judges, particularly in the Southern District, remain meaningfully more receptive to preliminary-injunction motions in restrictive-covenant cases than their state-court counterparts. The Commercial Division has trended employee-friendly; the federal trial courts less so.

Forum selection matters. If your contracts give the employer the right to choose forum, the federal court is often the better venue for an injunction motion. If your contracts are silent, expect the employee to remove or counter-file in state court at the first opportunity.

What this means for employers in 2026

1. Tighten your drafting now. Overbroad clauses are not just unenforceable — they're affirmatively risky, because severability protections are weakening.

2. Pick your battles. If you can only enforce one clause, enforce the employee non-solicit or the confidentiality clause. Save the non-compete fight for the case where you have facts.

3. Build the file. Document what the departing employee actually had access to, who they actually serviced, and what they actually contributed. Without specifics, you cannot meet the new irreparable-harm standard.

4. Consider garden leave. A 6-month garden leave is much more enforceable than a 12-month non-compete and is increasingly the market standard for senior commercial roles.

What this means for employees

You have more leverage than your contract says. Even if you signed an overbroad clause years ago, it may not survive judicial scrutiny in 2026. That doesn't mean ignore it — the cost of litigation is real and the uncertainty is real. But it does mean that when you're negotiating an exit, you should not assume the contract controls. It's a starting point, not a verdict.

Bottom line

The non-compete world is not what it was in 2020. The trend is toward narrower enforcement, sharper scrutiny, and more employee-favorable outcomes — even without a statutory ban. Employers who keep using 2019 contract templates are paying for protection they don't actually have. The fix is drafting, not litigation.