Starting a Startup While Still Employed: Moonlighting Clauses, Non-Competes, and Employer IP Claims

Most founders don't quit on day one. They build on nights and weekends while collecting a paycheck, and that is exactly when the most expensive ownership problems get created.

The story is familiar. An engineer has an idea, keeps the day job, and spends six months building a prototype after work. A co-founder joins. They incorporate, split the equity, and start talking to investors. The prototype works, the traction is real, and nobody has thought about the employment agreement the engineer signed three years ago.

That agreement matters more than almost any other document in the company's history. It can determine who owns the technology, whether the founder was even allowed to build it, and whether the company's most important asset belongs to the company or to someone else.

This post explains how an existing employer can reach into a startup, which clauses create the most risk, how state law limits those claims, and what founders should do before they write another line of code.

Why your employer may have a claim you never knew about

Employees often assume that what they do on their own time belongs to them. That is sometimes true, but it is a default that employment agreements are designed to change.

Invention assignment clauses can be broad. Most technology employers require employees to assign inventions to the company. Well-drafted clauses are limited to work done within the scope of employment. Many are not. They can reach anything related to the employer's business, anything developed using company resources, or anything conceived during the period of employment, whether or not it happened on company time.

Some clauses reach past your last day. "Holdover" or "trailer" provisions claim inventions created for a set period after the employee leaves, on the theory that the idea was probably formed while employed. Their enforceability varies by state, but they can turn a clean departure into a dispute.

Related business is a wide net. If the startup operates in the same space as the employer, or in an adjacent one the employer is exploring, the employer has a stronger argument that the work relates to its business. That is especially true in fast-moving fields like AI, fintech, and biotech, where a company's research roadmap can be broader than any one employee knows.

Company resources create evidence. Using a work laptop, work accounts, work cloud credits, proprietary code, or confidential information does more than breach policy. It gives the employer a factual basis to claim ownership and, in some cases, to allege trade secret misappropriation.

What state law protects, and what it doesn't

Several states limit how far an employer can go. California Labor Code section 2870 is the best known. It provides that an employer cannot require an employee to assign an invention developed entirely on the employee's own time, without the employer's equipment, supplies, facilities, or trade secret information, unless the invention relates to the employer's business or its actual or demonstrably anticipated research or development, or results from work performed for the employer. Other states, including Delaware, Washington, Illinois, Minnesota, and North Carolina, have similar statutes.

Founders tend to read these laws as a safe harbor. They are narrower than that.

The carve-outs have their own exceptions. The exception for inventions that relate to the employer's business or anticipated research is the one that matters most, and it is the one that applies to startups building in the same industry as their day job.

The statutes govern assignment, not everything else. They do not necessarily protect an employee from claims that they breached a duty of loyalty, misused confidential information, violated a moonlighting policy, or competed while still employed.

Which state's law applies is not always obvious. Employment agreements often include a governing-law clause, and the employee's work location, the employer's headquarters, and the founder's residence can all point in different directions.

The risks that stack up besides IP

IP ownership is the headline issue. It is rarely the only one.

Moonlighting and outside-activity policies. Many employers require approval for outside business activities, or prohibit them in the same field. Violating the policy can justify termination and can feed a later ownership claim.

Duty of loyalty. Even without a written non-compete, employees generally owe their employer a duty of loyalty while employed. That typically means not competing with the employer, not diverting business opportunities, and not using employer resources to build a rival. Preparing to compete is treated differently from actually competing, but the line is fact-specific and easy to cross.

Non-solicitation of coworkers and customers. Recruiting a colleague to be a co-founder, or approaching an employer's customers, is one of the most common triggers for a dispute. Non-solicitation obligations often survive employment, and a poorly timed conversation at the office can become Exhibit A.

Non-competes. Enforceability varies significantly by state. California generally refuses to enforce them; other states enforce reasonable ones. A founder who assumes a non-compete is void without checking the governing law is guessing.

Confidentiality obligations. Ideas, customer insights, technical approaches, and pricing information learned on the job can be protected even when no formal patent or copyright exists. The startup's roadmap should not rest on anything the founder learned only because of the job.

How this shows up in diligence

Investors do not just ask whether the company has signed IP assignments from its founders. They ask where the founders were working when the technology was created.

The founder's assignment can only transfer what the founder owns. If the employer has a valid claim to the technology, the founder cannot hand a clean title to the startup, no matter how carefully the company's own paperwork is drafted. The ownership gap sits one level above the company, which is why it often goes unnoticed until counsel reads the founder's employment agreement.

The representations in the financing documents make it personal. Standard Series A documents include representations that the company owns its IP and that no third party has asserted a claim. If a prior employer's claim exists and is not disclosed, founders face indemnification exposure.

Timing matters. A founder who left the job before any work began, with clean separation and no shared resources, looks very different in diligence from one who built the core product over a year while employed, using a work laptop for some of it.

Honesty is a requirement. Investors will ask founders to describe their employment history and any overlap with the company's technology. Understating the overlap is far more damaging than the overlap itself.

What founders should do before they build

Read your employment agreement and any IP policy now. Look specifically for the invention assignment clause, the definition of the employer's business, any outside-activity or moonlighting restriction, confidentiality terms, holdover provisions, non-solicitation language, and the governing-law clause. Do this before the first commit, not after the first term sheet.

Complete the prior inventions list carefully. Many agreements include an exhibit where employees list inventions they already own and want excluded. A thorough, specific list at the start of employment is evidence of what predated the job. One that is vague or left blank is hard to fix later.

Keep the startup completely separate. Use personal devices, personal accounts, personal cloud storage, and personal time. Do not copy code, documents, or data from the employer, and do not discuss the startup on company systems. Separation is not a guarantee, but it removes the easiest facts for an employer to rely on.

Consider asking for a written release. If the project is in a gray area, an explicit written waiver or confirmation from the employer that it claims no rights in the work can resolve the issue in the best possible way. Whether to ask is a strategic decision, because asking can alert the employer, and it should be made with counsel.

Document the timeline. Keep dated records of when the idea was conceived, when development began, what tools and resources were used, and when the founder left the employer. Contemporaneous records are far more persuasive than a recollection years later.

Do not recruit from your employer while employed. If a colleague is a natural co-founder, wait until your own obligations allow the conversation, and check non-solicitation terms first.

Plan the exit. Think through when to resign, what to return, what to certify on departure, and what the exit conversation will cover. Leaving cleanly is part of the ownership analysis.

Get a legal review before raising. A short review of the founder's employment agreement against the company's technology, early in the process, is far cheaper than discovering a conflict during diligence.

Frequently asked questions

Can my employer really own something I built on my own time?

Sometimes. If your agreement assigns inventions broadly and no state statute limits it, or if the invention falls within an exception such as relating to the employer's business, the employer may have a valid claim. Building on your own time with your own equipment helps, but it does not settle the question by itself.

Does California Labor Code section 2870 protect me?

It may, but only if your invention was developed entirely on your own time without employer resources or confidential information and does not relate to the employer's business or its actual or anticipated research. If your startup is in the same field as your employer, the statute may not help. Other states have comparable laws with their own wording.

Do I have to tell my employer I'm starting a company?

It depends on your agreement and policies. Some require disclosure or approval of outside business activities, and failing to comply can create separate problems. Others are silent. Whether and when to disclose is a strategic decision that should be made after reviewing the documents.

Is it safe to build the startup in the same industry as my day job?

It is the highest-risk version of this situation. The closer the startup is to the employer's business, the more likely an invention assignment clause, a duty of loyalty argument, or a confidentiality claim applies. It can be done, but it needs careful structuring and often a written release.

What if my startup idea came from a problem I saw at work?

The idea itself is generally not protectable, but confidential information you learned on the job is. The distinction between a general insight and proprietary information such as customer data, internal metrics, or non-public technical details is where many disputes begin. Build from what you can legitimately use, not from what you learned in confidence.

Can I fix this after the company is already formed?

Often yes, but it is harder and more expensive. Options include obtaining a release from the employer, rewriting or rebuilding the affected portions of the product, negotiating a license or assignment, and disclosing the issue to investors with a clear plan. Waiting until diligence leaves the founder with the least leverage.

Building on nights and weekends is a legitimate way to start a company, and many successful ones began that way. The difference between a founder who raises smoothly and one who stalls is rarely the idea. It is whether anyone checked the employment agreement before the work began. If you are building a startup while employed, or about to leave a job to do it, and want to confirm the technology belongs to the company, contact Ana Law to schedule a strategy session.