Commercializing Research-Institute Technology in China

A Hefei research institute has developed a new industrial sensing technology. Several researchers want to commercialize it through a startup that will raise venture capital. The institute owns some patents, one patent application was filed jointly with a university, critical manufacturing parameters remain confidential know-how, and several researchers expect to join the startup after incorporation. The founders describe the plan as a “technology spin-out.” Legally, the startup needs a much more precise package of rights. China’s Patent Law provides that inventions made in execution of an entity’s tasks or mainly using its material and technical conditions can constitute service inventions, with the right to apply for a patent belonging to the entity under the statutory framework.^[1] The 2025 Anti-Unfair Competition Law protects trade secrets only where the information is non-public, commercially valuable and subject to corresponding confidentiality measures.^[2] The Civil Code’s technology-contract framework and the Supreme People’s Court interpretation on technology-contract disputes govern transfer, licensing and the allocation of rights in development and technical-secret arrangements.^[3] The revised Company Law also permits certain non-cash assets, including intellectual property rights, to be used as capital contributions where they can be valued and lawfully transferred.^[4]

The specific problem#

The core transaction problem is therefore how to move enough enforceable technology rights into the startup for investors to finance it without accidentally stripping the institute of rights it must retain or leaving the startup dependent on undocumented personal knowledge.

Technology inventory and service-invention ownership#

The first diligence exercise should identify every legal and practical component of the technology. The inventory should cover patents, patent applications, software, technical drawings, datasets, prototypes, manufacturing processes, laboratory methods, test results, documentation and know-how. For each item, record the creator, current owner, date of creation, funding source, relevant employment relationship and whether the information has been disclosed publicly. This exercise frequently reveals that the “technology” is not one asset. A startup may be offered an exclusive license to three patents while its product actually depends on an unpublished calibration process known by two institute engineers. If that process is not licensed or transferred, the company does not control the technology required to manufacture its product. Patent certificates alone are therefore insufficient for venture diligence. The investor should understand what the patent claims cover and what unpatented knowledge is necessary to practice the invention commercially.

The institute should also identify third-party rights. Joint research with a university, government-funded project, supplier or corporate sponsor may have created co-ownership, consent rights or field restrictions. A commercialization plan built before those agreements are reviewed can promise rights the institute does not control alone. The inventory becomes the schedule to the transfer or license agreement and the baseline for future improvements. Article 6 of the Patent Law provides the statutory framework for service inventions created in execution of an entity’s tasks or mainly using its material and technical conditions.^[1] The exact ownership analysis should be based on the researchers’ roles, project documents and applicable agreements. The spin-out should not assume that inventors personally own a patent merely because their names appear as inventors. Inventorship and ownership are different legal concepts. For each patent or application, counsel should review: The main points are applicant and registered owner, inventors, employment status when the invention was completed, research project assignment, use of institute facilities or funding, and prior assignment agreements.

If an application remains in a researcher’s name even though the institute claims it as a service invention, that discrepancy should be resolved before venture financing. The institute should also review inventor reward and remuneration obligations under the Patent Law and applicable internal policies. A commercialization transaction can become contentious if researchers believe the institute is transferring valuable technology without recognizing their statutory or contractual interests. The startup’s financing documents should not contain a broad representation that “the founders own all core IP” where the real ownership sits with the institute. There is no single correct commercialization structure. Assignment. The institute transfers ownership of patents or other transferable IP to the startup. This gives investors a clean ownership story but may require valuation, approval and institutional procedures, especially where public research assets are involved.

Exclusive license. The institute retains ownership while granting the startup exclusive rights in an agreed field, territory or market. A durable exclusive license can be financeable if investors are protected against arbitrary termination and the startup can sublicense or transfer rights in future transactions where appropriate. IP contribution. Where legally and institutionally permissible, intellectual property can be contributed as non-cash capital under the Company Law if it is capable of valuation and lawful transfer.^[4] The parties need reliable valuation and completion of the relevant transfer formalities. The correct choice depends on institute policy, funding rules, strategic objectives and the startup’s exit plan. A venture-backed company expecting multiple financing rounds and an eventual IPO or sale generally needs rights that survive founder departures, management disputes and changes in share ownership. A short personal-use license revocable by the institute will usually be problematic. The commercialization agreement should therefore be reviewed from the perspective of the next investor, not only the first founders.

Assignment, licensing and intellectual-property contributions#

The 2025 Anti-Unfair Competition Law defines trade secrets as non-public commercial information with value that is subject to corresponding confidentiality measures.^[2] A spin-out cannot simply state that it receives “all know-how.” The agreement should identify trade-secret categories or secret points with enough precision to establish what is being shared while avoiding unnecessary public disclosure. Examples may include: The analysis turns on manufacturing tolerances, calibration methods, source code, process parameters, supplier specifications, and experimental datasets. The parties should specify who can access the information, whether it can be disclosed to investors or contract manufacturers, and what happens if the startup later develops an improved version. Confidentiality measures need to continue after transfer or licensing. The startup should implement access controls, confidentiality agreements, repository permissions and supplier protections. If the institute shares a highly valuable secret without any restrictions and the startup then circulates it widely, later enforcement can become harder.

The commercialization agreement should also distinguish institute trade secrets from general expertise carried in researchers’ skills. The startup cannot realistically “own” every piece of professional experience in an individual’s mind. The startup may commercialize the first-generation technology but create most enterprise value through later improvements. the agreement needs to define ownership of: The critical items are improvements created solely by startup employees, improvements created solely by institute researchers, joint improvements, improvements using shared facilities, and inventions arising under future sponsored research. The Supreme People’s Court technology-contract interpretation addresses rights and transfers in technology development and technical-secret arrangements and emphasizes the importance of contractual allocation.^[3] A vague clause saying that “future improvements belong to the party that creates them” may be insufficient when personnel work across both entities. The parties can create a field-based structure: the institute retains academic research and non-commercial fields, while the startup receives exclusive commercial rights within defined applications. Another model gives each party ownership of its own improvements but provides cross-licenses.

Investors should understand whether a competitor can obtain substantially the same next-generation technology directly from the institute. A spin-out whose exclusive rights apply only to yesterday’s patent can lose value quickly if tomorrow’s improvement remains outside the company. Several institute researchers may join the startup. The legal team should determine: The core diligence set covers whether they resign or maintain dual roles, confidentiality obligations, conflict-of-interest rules, ownership of new inventions, access to institute facilities, and sponsored research arrangements. the startup needs to not rely on the assumption that because a researcher becomes a founder, all of that researcher’s future work belongs automatically to the startup. If the person remains employed by the institute or uses institute resources, service-invention and contractual issues may continue. Conversely, the institute should not attempt to claim every invention created by a former employee after departure merely because it relates to the same field. Clear employment and research agreements reduce this ambiguity.

Investors should also identify whether one researcher is the only person capable of practicing critical know-how. If so, retention and knowledge-transfer planning are commercial as well as legal priorities. The technology package should be usable by the company even if one founder later leaves.

Trade secrets, improvements and researcher mobility#

Where the institute is publicly funded or manages state-related technology assets, commercialization may require internal or external approval, valuation, recordal or other procedures under the applicable institutional and state-asset framework. The startup and investor should not assume that an institute director’s commercial term sheet alone can transfer the rights. The closing checklist should identify: The most important elements are approving body, valuation requirement, technology-transfer procedure, conflict approvals, signature authority, and patent transfer or license recordal where applicable. Any condition that remains incomplete should be a true financing condition rather than an informal promise to “fix paperwork after closing.” the investor needs to also confirm that the institute’s representative had authority to sign the commercialization agreement. If the transaction is challenged years later, a complete approval file protects both the institute and the startup. The purpose is not bureaucratic formality. It is title certainty. Before financing, That gives the investor a basis to ask a harsh question: if relations between the institute and founders deteriorate tomorrow, can the startup continue selling its core product?

The answer depends on: The practical focus is on ownership or license duration, termination grounds, source-code and documentation access, sublicensing, improvements, facility dependence, and patent-maintenance obligations. An exclusive license terminable whenever the institute “believes cooperation is unsatisfactory” is not secure enough for institutional capital. Termination should be limited to defined material breaches, with cure periods where appropriate. If the institute retains responsibility for patent prosecution and maintenance, The startup can then receive notice and step-in or cooperation rights if the institute plans to abandon a key patent. Change-of-control clauses also matter. An institute may reasonably want to prevent transfer to a prohibited party, but a blanket termination on any financing or acquisition can undermine exit value. The agreement can instead balance public research interests with the startup’s need for financeable rights. Assume a Hefei research institute owns two Chinese patents and one pending application for a new industrial sensor. A joint university project produced the underlying algorithm, while process calibration remains undocumented know-how held by institute researchers.

A venture fund wants to invest RMB 80 million. A weak structure would give the startup a five-year “right to use institute technology” and state that founders will provide necessary know-how. A stronger structure would: The sequence is confirm institute ownership of service-invention patents; obtain the university’s consent or license for jointly controlled algorithm rights; grant the startup a long-term exclusive commercial license or assignment for defined fields; document and securely transfer the calibration know-how; allocate improvements and future sponsored research; complete institute approval and valuation; and put researcher employment and invention ownership on a clean basis. The investor can then diligence a coherent asset package rather than a relationship.

Institutional approvals, investor diligence and sponsored research#

Technology ownership defects become more expensive to fix later. A Series A investor may accept informal cooperation because the founders and institute still have a strong relationship. A later strategic investor or IPO intermediary will ask for documentary proof of title, exclusive rights, approvals and inventor obligations. The commercialization file should therefore preserve: The main points are patent records, approvals, valuation, transfer/license agreement, research contracts, employee assignments, trade-secret schedules, and improvement records. Any side letter modifying rights should be stored with the main agreement. The startup should also create internal IP governance. New inventions should be reviewed for patent filing, secrecy and ownership. Researchers moving between institute and company should have clear project documentation. The goal is to prevent the spin-out from recreating the same ownership ambiguity with its second generation of technology. Founders often ask for the broadest possible rights because they fear future dependence on the institute. Investors may encourage the same approach. That can create unnecessary valuation, approval and negotiation problems.

The company should identify which background patents, know-how and software are genuinely necessary for the planned products. Rights outside that field may be left with the institute. A focused license can be stronger than a vague license to “all institute technology.” It is easier to value, easier to approve and easier to monitor. It can also reduce conflict with other institute commercialization projects. The field definition should still be broad enough to cover foreseeable product evolution. If the startup begins with industrial sensors but expects to expand into medical sensing, the agreement should address whether that adjacent field is included, optioned or excluded. The investor should understand the boundary before pricing the company. A spin-out often continues to use institute laboratories after the initial technology transfer. That continuing research should not rely on the original license alone. A sponsored-research agreement should define: The analysis turns on project scope, funding, personnel, facilities, ownership of new results, publication, confidentiality, and patent filing.

The Civil Code technology-development framework and the Supreme People’s Court interpretation make contractual allocation particularly important where parties collaborate on development and technical secrets.^[3] the startup needs to avoid a situation where venture funds pay for research but the resulting IP automatically belongs entirely to the institute because the relationship was undocumented. The institute likewise needs protection against the startup claiming rights in unrelated academic work. Separating the initial technology transfer from future research makes both relationships clearer.

Data, software, enforcement and financing conditions#

Research projects increasingly produce datasets, models and software alongside physical inventions. The startup can then identify whether it needs rights to: The critical items are training data, experimental databases, source code, simulation models, and laboratory-control software. Ownership and transferability can differ from patent rights. Datasets may contain personal information, confidential third-party information or contractual restrictions. Software may include open-source components. The technology-transfer agreement should therefore contain separate schedules for software and data where they are material. An investor should not assume that a patent license includes source code or research data merely because those assets were used to develop the patented invention. If the institute retains patent ownership and licenses the startup exclusively, future infringement creates a governance question: who decides whether to sue? the agreement needs to address: The core diligence set covers notice of infringement, control of enforcement, allocation of cost, settlement authority, and recovery of damages. The startup may need enforcement to protect its market, while the institute may be cautious about litigation.

A financeable exclusive license should give the startup meaningful protection if the owner refuses to enforce. The same issue arises with patent maintenance and foreign filings. If the startup plans export markets, it should know who decides which jurisdictions receive patent protection and who pays. These provisions can materially affect enterprise value even though they receive less attention than the initial license fee.

Case analysis and future commercialization governance#

Not every diligence issue needs to block financing. A missing employee confidentiality acknowledgment can often be remediated after signing. Uncertain ownership of the core patent usually cannot. the investor needs to classify findings into: The most important elements are title-critical, license-critical, governance remediation, and ordinary post-closing cleanup. Title-critical issues should be conditions precedent. Operational cleanup can be handled through covenants with deadlines. This discipline prevents the financing from being delayed by minor documentation while ensuring the company does not receive capital before it controls the technology on which valuation depends. The legal report should explain why each item belongs in its category rather than present a flat list of “risks.” Before closing, That gives the investor a basis to be able to hand the technology schedule to new counsel or a future buyer and have them understand what the startup owns, what it licenses, what remains with the institute and how improvements are allocated. If the answer still depends on one founder explaining the history orally, the commercialization structure is not yet diligence-ready.

That standard is useful because people change. Researchers leave, fund managers rotate and institute leadership changes. Durable enterprise value depends on documents that survive those changes. The commercialization documents should also address foreign patent strategy where international sales are expected. A startup that receives only Chinese rights may discover that the institute controls patent filings in the United States, Europe or other key markets. The agreement can instead identify who decides foreign filings, who bears prosecution and maintenance cost, and whether the startup can step in if the institute declines protection in a market central to the business plan. That issue can materially affect later financing and export value.

Conclusion#

Research commercialization is not completed by placing patents into a term sheet. A venture-backed spin-out needs a legally coherent package of patents, trade secrets, improvement rights, researcher arrangements and institutional approvals. The Patent Law governs service-invention ownership,^[1] the 2025 Anti-Unfair Competition Law defines and protects trade secrets,^[2] the Civil Code and Supreme People’s Court technology-contract framework govern technology transfer and licensing,^[3] and the Company Law provides a basis for qualifying intellectual property to be contributed as non-cash capital.^[4] The final point is: finance the startup only after the technology rights have been separated from personal relationships and documented so the company can operate even if the institute, founders and investors later disagree.

Legal and regulatory sources#

^[1] Patent Law of the People’s Republic of China (2020 Revision), including Article 6 on service inventions: https://www.npc.gov.cn/npc/c2/c30834/202011/t20201119_308800.html ^[2] Anti-Unfair Competition Law of the People’s Republic of China (2025 Revision), especially Articles 10 and 39: https://www.npc.gov.cn/npc/c2/c30834/202506/t20250627_446247.html ^[3] Supreme People’s Court, Interpretation on Several Issues Concerning the Application of Law in the Trial of Technology Contract Dispute Cases: https://wb.flk.npc.gov.cn/sfjs/texthtml/84711b70b2114140bda3644cfa10bfe6.html ^[4] Company Law of the People’s Republic of China (2023 Revision), including the provisions on non-cash capital contributions: https://www.npc.gov.cn/npc/c2/c30834/202312/t20231229_433999.html

General legal information only; not legal advice for a specific technology-transfer transaction.