Who owns the patent? You, your business partner, your employer, or your contractor?
This post was originally published on Substack. If you would like to receive email alerts for new posts, I encourage you to visit and subscribe: https://open.substack.com/pub/stnagy/p/who-owns-the-patent-you-your-business
Your company designs a new surgical tool that addresses a real problem and stands to earn it millions. To build the prototype, you bring in a technician to assist your inventor. The inventor wants to add a safety sensor to alert a surgeon the moment the new tool punctures the wall of a body cavity, and the technician must redesign the sensor to make it work in the new tool.
Your tool launches, succeeds, and draws copycats almost immediately. You sue, and while the suit is pending, one infringer learns of the technician’s contribution, takes a retroactive license from him, and moves to correct inventorship to add him to the patent, making him a co-owner. Because a patent infringement suit cannot proceed without every co-owner, and the technician will not join, your case is dismissed.[1] The infringer stays on the market, and your share erodes.
That actually happened to Ethicon.[2] Ethicon and the inventor sued a competitor, U.S. Surgical, and lost, precisely because of this inventorship blunder. The technician, who was originally unnamed, was found to have contributed to two claims out of fifty-five in the patent, neither of which were actually asserted![3] Yet a co-inventor on even one claim is a co-owner of the entire patent.[4]
Invention ownership hygiene is easy to get right, but it is also easy to get horribly wrong. This piece discusses the default patent ownership rule (and its historical origins), the employer ownership “tiers” that can apply to employee inventions, common scenarios (moonlighting, joint founder inventions, contractors), and a chain-of-title checklist for owners and IP administrators.
God Save the King (but not his monopolies)
Since 1790, the U.S. patent law has operated on the premise that rights in an invention belong to the inventor, not the employer.[5] Unless there is an agreement to the contrary, it is settled law that “an employer does not have rights in an invention which is the original conception of the employee alone.”[6] This structure fundamentally contrasts with some other areas of IP law, like copyright, which utilizes a “work-made-for-hire” doctrine to legally deem the employer as the original creator and owner from the moment a copyrightable work is made by an employee, within the scope of her employment.[7]
At its roots, the patent ownership rule is fundamentally American. The framers drafted the authority for the U.S. patent system, contained in Article I, Section 8, Clause 8, to ensure Congress would secure exclusive patent rights to “Inventors.”[8] In doing so, the framers, wary of government-backed monopolies, broke with the English Crown’s habit of granting patents to court favorites for trades that already existed, a practice bad enough that Parliament itself had reined the Crown in with the Statute of Monopolies of 1624, carving out an exception only for “the true and first inventor” of something genuinely new.[9]
The patent ownership rule was solidified by the Supreme Court in the 1930s in United States v. Dubilier Condenser Corp., holding that unless an employee is specifically “hired to invent” a distinct item, a patent to an invention based on their labor remains the personal property of the employee. The employer’s payment buys, at most, a “shop right,” but not ownership.[10]
Ownership, Maybe, and Not Really
Patent assignments are the legal tool to convey ownership from an employee to their employer, but not all assignments are created equal. There are three common “flavors” that operate to transfer rights to an employer, but one is far more preferable than the other two. In order, they are:
Tier 1 — Express written assignments
A written patent assignment is by far the best option, and the only option that can be officially recorded against a patent at the patent office. Just take care to ensure your written assignments actually accomplish what you need them to do.
Often, a patent assignment is not really an assignment, but simply a promise to assign the patent sometime in the future. Statements like “employee agrees to assign” or “employee shall assign” are obligations (or promises) to assign a patent; these statements do not assign a patent (or its underlying application) on their own.
And, until an assignment actually occurs, a promise like these conveys equitable rights to the employer, but not legal title. Arachnid, Inc. v. Merit Industries, Inc. is a clean illustration: a consulting agreement saying inventions “shall be the property of” the client and that rights “will be assigned” gave the client equitable title only, and therefore no standing to recover for infringement occurring before it obtained record title.[11]
In your assignments, you need stronger language, e.g.: “I hereby assign.” This recitation transfers legal title automatically, by operation of law, the moment the invention comes into existence, and no further paperwork is required. It is, unfortunately, one of those “technical drafting trap[s] for the unwary,” as observed by Justice Breyer. [12]
Tier 2 — “Hired to Invent”
Absent an express assignment, someone “employed to make an invention” who succeeds is legally bound to assign the resulting patent to their employer . . . maybe.[13] The obligation is a kind of implied-in-fact contract, which applies when an employer specifically hires an employee to exercise inventive faculties on a particular problem.[14]
While it sounds nice in theory, the doctrine is cabined in important ways that make it unwise to rely on.
First, it is hard to predict with certainty whether the “hired to invent” doctrine will apply. For example, an employee is generally not hired to “make an invention” simply because their job description involves research activities; and even full-time R&D employment may not be enough on its own.[15] Indeed, the wording of the employment agreement (may or) may not be the deciding factor.[16]
Second, whether an implied-in-fact contract exists to support the hired-to-invent argument depends on state law and a judge’s interpretation of your agreement.[17] So, if a dispute arises, you may not know whether you have any rights until after discovery, motions practice, and possibly a trial. These kinds of disputes can become incredibly expensive and time consuming; if a competitor takes a license from your employee like in the Ethicon case above, the competitor may decide it is in their interest to make that litigation as painful and expensive as possible.
Tier 3 — Shop Right
Sometimes employees have not signed assignment agreements and are not subject to the “hired to invent” doctrine. What happens if such an employee invents using an employer’s own time and materials? In these circumstances, the common law has developed a limited “shop right” for the employer, to allow it to continue to use the technology its own resources paid to develop. Such a “shop right” is essentially a non-exclusive, royalty-free license to practice the invention.[18] But true ownership of the patent stays with the employee.
For many companies, and especially for startups, a shop right has little value, or is completely worthless. A license to use a company’s core technology, with the actual patent sitting with the employee, means the company cannot stop a competitor from licensing the same invention from the employee. If the competitor decides not to take a license, the company also has no ability to force the employee to sue the competitor to make them stop.
Moreover, a shop right also is generally treated as personal to the employer who earned it, meaning it is not licensable, not assignable, and not saleable as a standalone asset, though it is usually understood to pass with a sale of the entire business.[19] Any competent diligence team will flag a shop right as not belonging to the company.
Practical scenarios where patent ownership can become complicated
With an understanding how patent rights can be transferred from an employee to employer, let’s consider three of the more common scenarios in which companies who are aware of the basic rules can still make critical ownership mistakes:
Scenario 1 — Moonlighting: can your employer claim your startup’s intellectual property?
It is not uncommon for companies to launch with humble beginnings while their founders are still in school or employed at other jobs. Bill Gates and Paul Allen started Microsoft while at college and working, Larry Page and Sergey Brin built Google as student researchers, and Steve Jobs worked at Atari while building Apple.
Most founders never think about documents they signed with their academic department or at their last job, but the truth is many standard employment and university contracts contain broad invention-assignment clauses as part of initial onboarding. These can be easily forgotten as they usually have no occasion for renewal. But knowing the scope of IP obligations to your current or former employer is obviously critical. Founders, especially those working at the technical frontier, should be planning their IP position before the company is even incorporated, so it is critical to review paperwork from any employer or university that the founder had a relationship with during the early days of their new company.
Sometimes the question is how far an assignment clause extends; other times, the question is how far the clause is legally allowed to reach. On the former, before filing any patent application, founders should schedule a discussion with a patent lawyer (which they should be doing anyway before filing any application).
On the latter, IP assignment clauses can extend far, but there are public policy rationales that have led at least some states to place limits. Ten have passed statutes that carve genuine outside-of-work inventions out of the scope of employee invention assignment agreements. Although it is not the majority a founder might hope for, California is among them.
In California, for example, Cal. Lab. Code § 2870(a) provides that an assignment clause does not reach an invention the employee developed “entirely on his or her own time without using the employer’s equipment, supplies, facilities, or trade secret information,” with two exceptions: inventions that (1) relate, at the time of conception or reduction to practice, to the employer’s business or actual or demonstrably anticipated research or development, or (2) result from any work the employee performed for the employer.[20]
Any clause that tries to reach further is unenforceable in California as against the state’s public policy.[21] Yes, the exceptions (1) and (2) are fairly broad. For example, a founder who built a side project in a niche their employer might occupy cannot avoid assignment by using their own equipment on the weekend. Yet California’s statute nevertheless does place at least some limits on how far companies may reach.
Other states have similar protective statutes on the books: for example, Washington (RCW 49.44.140), Delaware (Del. Code tit. 19, § 805), Illinois (765 ILCS 1060/2), Kansas (Kan. Stat. § 44-130), Minnesota (Minn. Stat. § 181.78), New Jersey (N.J. Rev. Stat. § 34:1B-265), New York (N.Y. Lab. Law § 203-f), North Carolina (N.C. Gen. Stat. § 66-57.1), and Utah (Utah Code § 34-39-3).
Thus, to avoid the mistake of losing IP to a former employer or university, founders should be sure to check their agreements, knowing that state law may (or may not) play a role in an IP strategy going forward.
Scenario 2 — Your co-founder: A joint invention can be a joint checking account.
Most modern patents contain at least two inventors. Solo inventorship has been declining for some time, according to Patently-O: “the average number of inventors per utility patent has reached 3.2 in 2024, nearly double the 1.7 inventors per patent seen in 1976.”[22]
As Ethicon (hopefully) showed, absent an agreement otherwise, a joint inventor of a patent is, by default, also a joint owner. Joint ownership of a patent is like holding your own independent keys to the gold mine. Each owner of a patent may make, use, offer to sell, sell, or import the invention “without the consent of and without accounting to the other owners.”[23] They both have the keys to unlock the asset, and neither needs the other’s permission to do almost anything with it, including licensing it.[24]
Practically speaking, if joint owners of a patent do not agree to cooperate, one can completely undermine the value of the asset for a business they share. Take, for example, a company in discussions with a potential acquirer that is interested in the company’s technology. If the co-founder / co-inventor has no overriding fiduciary obligation, they could circumvent the company and deal directly with the acquirer, making a deal to license core IP rights directly. The acquirer would have no need to make a deal with the company.
Or, for example, a copycat facing patent litigation exposure based on your company’s IP. The co-founder / co-inventor could again circumvent the company and deal directly with the copycat, making a deal to refuse to join any enforcement action, thereby, unilaterally vetoing it, to the benefit of the copycat.
So, do not rely on a handshake with your co-inventor. Formalize ownership by executing an assignment of invention rights to the company, so NewCo (not the founders) holds the patent, and the founders hold equity instead of a fractional, independently-licensable interest in the underlying IP.
Scenario 3 — Your vendor: is IP a “deliverable” (or is it something else)?
Contractors and vendors are ubiquitous in modern business. They specialize in providing a certain service or good, so smart founders often make use of them to reduce costs and investment in areas outside the company’s core competency. But there are a couple of traps to avoid when working with them, to ensure any IP developed is owned by the company, not the contractor or vendor.
First, vendor agreements will often promise everything without actually assigning any intellectual property. A statement-of-work (SOW) or master services agreement (MSA) might say something like: “Client shall own all deliverables. ” To a non-lawyer, it reads like ownership of all work product, including IP, and a vendor may well contend that it transfers some IP, such as copyright in code. But this kind of loose language does nothing for patent rights unless it contains an actual assignment of inventions, which is a different legal act from owning work product.[25] Nor should you assume the copyright half is safe: a contractor’s work is a “work made for hire” only if the parties signed a written agreement saying so and the work falls into one of nine enumerated statutory categories, and software code is not one of them.[26] Bottom line: your vendor agreements must have an IP assignment clause.
Second, vendors often hire subcontractors to assist with duties set in the main vendor agreement. These subcontractors (another firm, a freelancer, an offshore team) do not have a relationship with you. Their only obligation is to the vendor who hired them. Whatever careful assignment language a company received from the vendor is worthless if the actual inventor is two layers down in a subcontractor who never agreed to assign rights. In other words, ownership can only flow down a chain that exists, and your vendor agreements must have a flow-down requirement binding every subcontractor to the same assignment obligations, so the chain of title has no gap.
A chain-of-title checklist to run right now
The commentary and scenarios above point towards one critical IP ownership question your company should be able to answer:
For every individual helping to develop your technology (founders, employees, contractors, vendors), does your company have a properly-worded assignment showing chain-of-title?
If you are not sure the answer is a resounding “yes,” here are the steps you should be taking now:
For every founder: Founders should have assignment agreements transferring ownership of inventions to their company. Make sure they include any pre-incorporation IP; it is easy to overlook. Double-check your prior agreements with employers/universities to ensure no third party might have an ownership claim in your IP.
For every employee. Employees should execute a Proprietary Information and Inventions Assignment Agreement (PIIAA) during onboarding that uses “hereby assigns” language (not “agrees to assign”). Some state laws require notices around independent inventions an employee is allowed to retain. Make sure the agreement also includes immunity provisions required by the Defend Trade Secrets Act.
For every contractor and vendor. Contractors and vendors must expressly assign intellectual property, not just state the client “owns all deliverables.” Agreements should also include invention disclosure obligations, flow-down terms requiring subcontractors to assign inventions up the chain, and a power-of-attorney clause that lets you step in if needed.
And once you have secured rights from your personnel and partners, make a habit of engaging in patent ownership hygiene in at least these two scenarios:
In every application. File confirmatory assignment(s) executed and recorded at the USPTO within three months (best practice). Assignments recorded later can be void against any bona fide purchaser for value.
Before every diligence round. Double-check that you can demonstrate an unbroken, signed chain from every named inventor to the company without chasing anyone down. This is a routine document request during diligence, and a broken ownership chain can become a major headache, possibly derailing discussions.
An ounce of prevention is worth a pound of cure. Likewise, just a little attention to IP ownership at the beginning, when all involved are optimistic and agreeable, can avoid significant investments of time and money down the road if something goes wrong.
Questions? Comments? Reach out to me directly (seriously)!
Related reading
Talking to investors before you have filed carries its own disclosure risk; see “Can I talk to investors or demo my product before I file a patent?”
And one layer upstream of everything in this piece: the ownership analysis does not matter if the “inventor” does not qualify; read the deep dive on how AI impacts patentability in “An AI agent found your lead, who owns the patent?” for the threshold question that should be answered first.
If you made it this far, I owe you a beer.
Grab one with me in Washington, DC.
Author Note
Stephen G. Nagy is a patent attorney and engineer at Strain PLLC. This article is for general informational and educational purposes only, is not legal advice, and does not create an attorney-client relationship. It reflects the author’s views, not necessarily those of Strain PLLC. Legal authorities change; verify currency before relying on anything here. For advice on your specific situation, consult qualified counsel.
This article is my own work:
Footnotes
[1] Ethicon, Inc. v. U.S. Surgical Corp., 135 F.3d 1456, 1467–68 (Fed. Cir. 1998) (“An action for infringement must join as plaintiffs all co-owners”; “as a matter of substantive patent law, all co-owners must ordinarily consent to join as plaintiffs in an infringement suit”).
[2] Id. at 1459 (“Yoon thereafter granted an exclusive license under this patent to Ethicon.”).
[3] Id. at 1459, 1465 (patent issued “with fifty-five claims”; district court found Choi contributed to the subject matter of claims 33 and 47). Ethicon sued on claims 34 and 50.
[4] Id. at 1465–66 (“Thus, a joint inventor as to even one claim enjoys a presumption of ownership in the entire patent.”).
[5] Bd. of Trs. of the Leland Stanford Junior Univ. v. Roche Molecular Sys., Inc., 563 U.S. 776, 780 (2011) (Roberts, C.J.) (“Since 1790, the patent law has operated on the premise that rights in an invention belong to the inventor.”).
[6] Id. at 786 (quoting United States v. Dubilier Condenser Corp., 289 U.S. 178, 189 (1933)).
[7] 17 U.S.C. § 201(b). For independent contractors, the doctrine does not operate automatically: § 101 treats a specially ordered or commissioned work as made for hire only if the parties sign a written agreement to that effect and the work falls within one of nine enumerated categories (a list that does not include computer software).
[8] U.S. Const. art. I, § 8, cl. 8 (Congress may “promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries”).
[9] Statute of Monopolies 1623, 21 Jac. 1, c. 3, § VI (excepting from the Act’s general prohibition any grant “of the sole working or makinge of any manner of new Manufactures within this Realme, to the true and first Inventor and Inventors of such Manufactures”); See also, United States v. Dubilier Condenser Corp., 289 U.S. 178, 189 (1933) (“The grant of letters patent is not, as in England, a matter of grace or favor, so that conditions may be annexed at the pleasure of the executive.”).
[10] 289 U.S. 178, 187–89 (1933) (“One employed to make an invention, who succeeds, during his term of service, in accomplishing that task, is bound to assign to his employer any patent obtained”; and, where the employee used “his master’s time, facilities and materials,” the employer gets a non-exclusive right to practice, but “has no equity to demand a conveyance of the invention.”).
[11] Arachnid, Inc. v. Merit Indus., Inc., 939 F.2d 1574, 1576, 1580–81 (Fed. Cir. 1991); FilmTec Corp. v. Allied-Signal, Inc., 939 F.2d 1568, 1572 (Fed. Cir. 1991). On the standing point, see Arachnid, 939 F.2d at 1579 (“one seeking to recover money damages for infringement of a United States patent (an action ‘at law’) must have held the legal title to the patent during the time of the infringement”); id. at 1580 (equitable title holder may still obtain “injunction, accounting, declaration of trust, or other forms of equitable relief”).
[12] Stanford v. Roche, 563 U.S. at 800-01 (Breyer, J., dissenting).
[13] Dubilier, 289 U.S. at 187; Standard Parts Co. v. Peck, 264 U.S. 52 (1924).
[14] Teets v. Chromalloy Gas Turbine Corp., 83 F.3d 403, 407 (Fed. Cir. 1996) (“[A] court must examine the employment relationship at the time of the inventive work to determine if the parties entered an implied-in-fact contract to assign patent rights.”).
[15] Dubilier, 289 U.S. at 187 (“[I]f the employment be general, albeit it cover a field of labor and effort in the performance of which the employee conceived the invention for which he obtained a patent, the contract is not so broadly construed as to require an assignment of the patent.”); see also Banks v. Unisys Corp., 228 F.3d 1357, 1359–60 (Fed. Cir. 2000) (vacating summary judgment for the employer where the employee had repeatedly refused to sign the company’s assignment form).
[16] Teets, 83 F.3d at 405, 407–08 (Teets was “an employee at will and had no written employment contract addressing ownership of inventive work”; he “spent at least 70% of his time on the GE90 Project”; “DRB has paid and continues to pay for the prosecution of a patent application for the HFP”; and Teets “stated ‘DRB devised or developed’ the HFP”).
[17] Teets, 83 F.3d at 407 (“[S]tate contract principles provide the rules for identifying and enforcing implied-in-fact contracts.”).
[18] Dubilier, 289 U.S. at 188–89; McElmurry v. Ark. Power & Light Co., 995 F.2d 1576, 1580–81 (Fed. Cir. 1993) (a shop right is “a right that is created at common law, when the circumstances demand it, under principles of equity and fairness, entitling an employer to use without charge an invention patented by one or more of its employees without liability for infringement”).
[19] Hapgood v. Hewitt, 119 U.S. 226, 234 (1886) (“Whatever license resulted to the Missouri corporation, from the facts of the case, to use the invention, was one confined to that corporation, and not assignable by it. … As to any implied license to the assignor, it could not pass to the assignee.”). Note that the successor in Hapgood was a corporation organized by the same stockholders, which the Court agreed “may naturally have succeeded to the business.” Yet the license still did not pass.
[20] Cal. Lab. Code § 2870(a). Section 2872 separately requires an employer whose agreement contains an assignment clause to give the employee written notice that the agreement does not reach inventions qualifying under § 2870, and puts the burden of proof on the employee claiming the benefit of the section.
[21] Cal. Lab. Code § 2870(b) (such a provision “is against the public policy of this state and is unenforceable”); see also id. § 2871 (referring to a provision “made void and unenforceable by Section 2870”).
[22] Dennis Crouch, The Team-Based Reality of Modern Innovation: Average Patent Now Lists More Than Three Inventors,Patently-O (Nov. 18, 2024), https://patentlyo.com/patent/2024/11/reality-innovation-inventors.html.
[23] 35 U.S.C. § 262.
[24] A co-owner’s license operates prospectively only; it cannot release another co-owner’s claim to accrued damages for past infringement. Ethicon, 135 F.3d at 1467 (following Schering Corp. v. Roussel-UCLAF SA, 104 F.3d 341, 345 (Fed. Cir. 1997)). And the refusal-to-join veto has two exceptions: a patentee who has granted an exclusive license must permit the licensee to sue in his name, and co-owners may contract away the right to refuse. Id. at 1468 n.9.
[25] In Arachnid, a consulting agreement providing that inventions “shall be the property of CLIENT” and that “all rights thereto will be assigned” was held to give the client equitable title only, leaving it without standing to recover for infringement occurring before it obtained record title. 939 F.2d at 1576, 1580–81.
[26] 17 U.S.C. § 101 (definition of “work made for hire”).
