Trade secrets in Uzbekistan: protecting know-how and client lists

A year ago the owner of a distribution company came to us. His commercial director had resigned, set up his own firm, and within three months had peeled off the eight largest clients — same prices, same payment terms, same decision-maker contacts. In effect he had walked out with the most valuable asset the business owned: a client base complete with terms. The owner wanted to sue for disclosure of a trade secret. We asked one question: "Did you have a trade-secret regime in place — an order, a schedule of protected information, a signed acknowledgement from staff?" He did not. Which means there was, legally, no secret at all: there was nothing to walk off with, and the claim was lost before it was filed. This article is about how not to end up in that room.

What the law actually treats as a trade secret

Uzbekistan has a dedicated Law on Trade Secrets. It does not protect "anything confidential" — it protects information that meets three conditions at the same time:

1. it has actual or potential commercial value precisely because it is unknown to third parties;
2. there is no lawful free access to it;
3. its holder has put a trade-secret regime in place over it.

Remove any one of the three and the protection is gone. Information can be worth millions, but if it is freely accessible or you never set up the regime, the law does not cover it.

What usually qualifies in practice: a client base with terms (prices, discounts, payment delays, contact people), production recipes and technology not disclosed in a patent, cost and margin calculations, the terms of supplier contracts, methodologies, drawings, the core source code of a product, pre-launch marketing plans.

What can never be a secret — however much you might want it to be. The law expressly excludes a set of information from the regime: constituent documents, headcount and working-conditions data, information a company is legally obliged to disclose (filings to state bodies, licences, pollution and violation data). You cannot classify your accounts to hide them from the tax authority or your staff.

The trade-secret regime: four steps, without which there is no secret

This is where founders go wrong most. "Everyone has signed an NDA" is not yet a trade-secret regime. An NDA is one of four elements, and on its own it will not persuade a court. The regime is a documented set of measures. Four steps at minimum:

1. The schedule. Approve, by internal order, a list of the information that constitutes a trade secret. Not "all company information" (a court throws out a schedule that vague) but specifics: "the client base in the CRM", "the recipe for product X", "the cost calculation". The more precise, the stronger in court.
2. Access. Restrict the circle of people admitted to the secret and keep a record of who is cleared for what. If the "secret" is reachable by the whole company through a shared folder, it is not a secret.
3. Contracts. Put a confidentiality clause into employment contracts and into contracts with counterparties, backed by a standalone NDA. With an employee, write in the duty to keep the secret after they leave too — for whatever period you state in the agreement.
4. Marking. Stamp physical media and documents with a "Trade Secret" label naming the holder. Without the mark it is almost impossible to prove a person knew a particular document was confidential.

It sounds like paperwork, but those four documents are exactly what turns "we all understood it was secret" into "here is the order, here is the schedule, here is the employee's signature acknowledging it." The first does not work in court; the second does.

And one feature that sets a secret apart from a trademark or a patent: setting up the regime costs no state fee and requires no registration at the IP Center. Registering a trademark costs around 1,200,000 UZS per class, and a patent application discloses the substance of the solution in exchange for a 20-year monopoly — whereas a trade-secret regime costs not a single som in fees and discloses nothing. You pay only in internal discipline. The flip side: nobody but you will set the regime up or check it — until it is too late.

An employee leaves for a competitor: what you can actually do

The most common leak is not a hacker — it is a departing employee. The law splits this into two periods.

While the person is employed, the duty of confidentiality follows from the regime and the employment contract. Disclosure is a disciplinary breach, up to and including dismissal, and at the same time grounds to claim damages.

After they leave, the duty survives exactly as far as you wrote it into the non-disclosure agreement. If the contract says nothing about the post-employment period, the former employee is free. That is why the clause "I undertake not to disclose for N years after the employment ends" is not a formality — it is the one thing left in your hand once the person has already gone.

What to do on day one, when you learn of a leak:

• secure the evidence (client statements, correspondence, matching terms, screenshots) — proof is gathered immediately, after which it disappears;
• confirm the regime was in place before the leak, and pull the documents (the order, the schedule, the signed acknowledgement);
• send a pre-action letter demanding that use stop and damages be paid;
• if the scale is serious, prepare a claim.

Liability for disclosure can be disciplinary, financial (damages, including lost profits) and, in serious cases, administrative and criminal. But the foundation of all of it is the regime. No regime, no liability — because there was nothing to breach. Where an employee's code, design or other authored work is involved, another layer of rules applies — we covered it in our piece on works made for hire.

Secret or patent: when to choose which

This decision is made once, and often wrongly. The logic runs like this.

A patent discloses the substance of the solution to the whole world in exchange for a monopoly of at most 20 years, and only in the countries where you obtained it. Once the term expires, the solution becomes public domain. In return, a patent protects you even against someone who arrived at the same thing independently.

A trade secret discloses nothing and lasts as long as you like — but only for as long as it stays secret. It has two weaknesses: it does not protect you against someone who genuinely invented the same thing themselves or lawfully took your product apart (reverse engineering), and it collapses the moment of the first leak.

The classic reference point is the Coca-Cola formula: protected as a secret since 1886, almost 140 years. A patent would have given at most 20 years and would have expired long ago, leaving the formula in the public domain. But that only works because the formula cannot be reliably reconstructed from the finished drink.

The practical rule: if the solution is visible in the product and easy to reproduce by taking it apart, patent it before it is too late (the application must outrun the disclosure). If the solution simply cannot be seen from the outside — an algorithm, a recipe, a method inside a closed process — the secret regime is both cheaper and more durable. Software code carries its own nuances, which we set out in our piece on protecting software.

What you must prove in court — and why claims fail

To win a disclosure dispute, the claimant has to prove three things:

1. the information met the marks of a trade secret (value plus restricted access);
2. the regime was in place — and in place before the leak happened;
3. the defendant either accessed it unlawfully or breached a duty they had taken on.

Claims almost always fail on the second point. The company tries to set up the regime after the fact, once the leak has happened — and the court sees it. Or the regime exists on paper, but the employee never signed an acknowledgement, the schedule reads "any company information", and there is no mark on the documents. Every such gap is a reason to refuse.

Even after winning on liability, the claimant runs into quantifying the loss: the profit lost from departed clients is hard to calculate to the precision a court accepts. So the realistic strategy is not only recovery but an injunction against further use, plus a record of the breach for future disputes.

The conclusion is unwelcome but honest: a trade-secret case is won not in the courtroom but a year earlier — on the day you signed the order putting the regime in place. If you do not have that order, start there, not with a claim. When it comes to litigation, those documents are what decide the outcome.

In short

• A trade secret protects what you cannot register: the client base, recipes, supplier terms, know-how.
• Information is protected only where three marks are present: value, restricted access and an active regime. No regime, no secret.
• The regime is four steps: a schedule, restricted access, a contract clause plus an NDA, and a mark on the media. One NDA does not count.
• The regime carries no state fee and no registration at the IP Center — but nobody other than you will set it up.
• Claims fail on proving the regime existed before the leak; setting it up after the fact is pointless.

FAQ

Do I have to register a trade secret with a state body? No. Unlike a trademark or a patent, a trade secret is not registered. It arises the moment the holder puts a trade-secret regime in place. There is no state fee — but equally, nobody is obliged to check that your regime is done correctly.

Is signing an NDA with staff enough? No. An NDA is only one element of the regime. Without an approved schedule of information, restricted access and a mark on the documents, a court may not accept that a regime existed at all, and then the NDA hangs in mid-air.

Can I stop a former employee from working for a competitor? You cannot ban the work itself — that is beyond the scope of a trade secret. But you can prohibit them from using and disclosing specific protected information, provided the regime was in place and the agreement set out a duty for the post-employment period.

What if there was no regime and the leak has already happened? You cannot back-date the regime — a court will spot it. General tools remain (breach of the employment contract, unfair competition), but they are weaker. The real lesson is to set the regime up before you need it.

How does a secret differ from a patent? A patent discloses the solution and protects it for 20 years in chosen countries, including against independent inventors. A secret discloses nothing and lasts indefinitely, but does not protect against honest reverse engineering and collapses on a leak.

Can information be a secret if the whole company can reach it? In practice, no. One of the marks is restricted access. If the "secret" sits in a shared folder open to everyone, a court will find that free access existed and refuse protection.

What liability does disclosure carry? Depending on severity — disciplinary (up to dismissal), financial (damages and lost profits) and, in serious cases, administrative and criminal. But all of it applies only where the regime was in place.

A trade secret is the only kind of intellectual property you create and lose entirely on your own, with no involvement from the state. The IP Center does not grant it and a court does not strip it — it holds up on exactly the discipline you imposed inside the company. The most valuable asset a business owns usually cannot be patented or registered as a mark. It can only be locked down with a regime — and that needs doing today, not on the day your commercial director has already handed in their notice.