Submitted by Lorena A. Barba, Faculty Director, GW Open Source Program Office (OSPO).
I offered these comments in my capacity as Faculty Director of the OSPO, drawing on our work supporting research software and open-source scholarship at GW.
These comments address the proposed policy from the perspective of research software and open scholarship. Page references refer to the 13-page consultation package, including its procedures and comparison document.
Research software is scholarly work in its own right. It embodies intellectual contributions, expresses research methods, and enables others to examine, reproduce, and extend scientific results. Its scholarly character does not depend on being attached to an article, serving an instructional purpose, or lacking commercial potential. University policy should recognize this explicitly and protect the academic freedom of principal investigators and their research teams to develop and disseminate software. It should also advance the public benefit that comes from access, reuse, collaboration, and continued research.
The draft provides a useful foundation. Its opening affirms GW's commitment to serving society and disseminating knowledge, and states that the university does not pursue knowledge primarily for commercial purposes. Its exclusion of copyrightable works, subject to an exception for patentable software, appears to improve the position of ordinary nonpatentable research software. The student examples, review targets, explanations when TCO declines protection, abandonment notice, and conflict-recusal provisions are also welcome.
The requested revisions below would make these commitments reliable in practice. The central concern is that inconsistent definitions, broad assignment language, and a process oriented toward protectability and marketability leave researchers uncertain about their authority to share scholarly software.
1. Foundational Principle: Software is Scholarship
Research software remains scholarship when it implements a patentable invention. Open-source dissemination can advance public benefit by enabling scrutiny, reproducibility, reuse, and collaborative development. Academic freedom and the PI's and research team's direction of their scholarly work should therefore inform how this policy governs covered inventions, alongside public benefit and applicable obligations. Commercial potential alone should not displace those considerations.
We are not proposing to reallocate copyright ownership through this policy. Ordinary nonpatentable software should remain governed by the Copyright Policy. The requests below concern the boundary between the two policies and the treatment of software-related inventions within this policy's scope.
2. Resolve the Software Scope Contradiction and Separate Rights
Pages 1 and 5 use different tests. The summary excepts “patentable software” from the copyright exclusion, while the ITRM definition excludes “unpatented copyrightable materials.” A potentially patentable software implementation can be unpatented. The two provisions therefore do not give researchers a consistent answer about coverage before publication or release.
Please distinguish patent rights in an invention from copyright in the code implementing it. These rights can coexist, but possible patentability should not automatically determine copyright ownership or move an entire codebase outside the Copyright Policy.
Proposed language:
“Copyright ownership in research software is governed by the Copyright Policy. This policy governs rights in covered inventions implemented through software and does not, by reason of patentability alone, determine ownership of copyright in the implementing code.”
The policy should also identify how a genuine coverage question is resolved, on what criteria, within what time, and what the researcher may do while the question is pending. Correcting “unpatented” to “nonpatentable” alone would not resolve these questions. Routine scholarly software should not require a universal patentability clearance before release.
3. Make Public Benefit an Operative Standard
The policy's public-interest commitments should guide its operative provisions. Page 8 identifies protectability and marketability as substantive review criteria. Those criteria alone do not adequately address the scholarly and public benefits of disseminating software that implements a covered invention.
Open-source dissemination should be recognized as a successful research outcome, including when no licensing revenue is anticipated. Relevant considerations include reproducibility, broad access, reuse, interoperability, education, community maintenance, and further discovery. These benefits should inform any institutional decision concerning protection, licensing, restrictions on dissemination, or return of rights.
The return provision on page 4 already invokes public benefit. That criterion should apply throughout the policy, including ordinary licensing decisions. Where an institutional restriction would depart from the research team's dissemination plan, GW should explain its specific justification, expected public benefit, and duration. Commercialization can advance the mission, but expected revenue should not displace academic freedom or the public value of sharing.
4. Provide a Clear Route for Open-Source Releases and Contributions
The package contains no explicit route for routine open-source releases or contributions. This is an omission that should be addressed, rather than left for researchers to interpret across overlapping policies.
Please affirm that creator-owned scholarly software outside this policy's scope does not require TCO commercialization clearance solely because it is software or was developed using ordinary scholarly resources. For software implementing covered inventions, where university-owned rights require authorization, provide advance authorization for defined projects, repositories, licenses, or classes of contributions, so that routine patches and releases do not require repeated review.
Guidance should accommodate development in public, contributions to existing projects, third-party licenses, and contributor license agreements. It should distinguish invention disclosure from authorization to distribute code and identify who may sign agreements covering university-owned rights.
The university should also implement open-source commitments retained in accepted sponsored-project terms. Researchers should be able to plan dissemination before work begins and rely on a documented authorization process. Such commitments do not themselves transfer ownership or authorize licensing of another party's rights.
5. Narrow and Clarify Ownership and Assignment Language
Pages 1–2 claim all rights in covered ITRM made with University Support and state that contributors hereby assign those rights effective from first use of that support, regardless of whether they execute a separate assignment. Page 2 also requires a signed invention assignment. The policy should explain the relationship among these provisions and identify the rights each concerns.
University Support, defined on page 6 to include time, facilities, materials, and nonpublic information, is not the Copyright Policy's Substantial Use test. Please state clearly that this definition does not override the copyright exclusion or replace the Copyright Policy's protections for ordinary scholarship. Routine research activity should not become a basis for claiming copyright in scholarly code through an uncertain ITRM classification.
Please clarify that these assignment provisions do not alter existing ownership or license terms for pre-existing code, or purport to transfer rights contributors do not hold. The revision should specify its effective date and treatment of ongoing projects and previously released software. The change from assignment that GW may require after evaluation to purported present assignment warrants explicit explanation; it is not merely a change in administrative wording.
6. Restore Bounded Publication Delays and Set Usable Decision Times
Pages 2–3, 5, and 8 need clearer timing safeguards. The current policy normally permits a sponsor patent-filing publication delay of up to 60 days. The draft retains timely-publication language but removes that numerical protection. Please restore it, define the triggering event, and specify any narrowly justified extension process.
The procedures target completeness review within 30 days and substantive review within 60 days after disclosure completion. These are not guaranteed release deadlines and do not resolve urgent needs associated with a paper, dissertation, conference, software release, or public development process.
Please establish expedited handling for time-sensitive scholarship, written decisions identifying any restriction, and escalation when deadlines are missed. A classification or commercialization review should not create an indefinite publication hold. The policy should distinguish these internal service deadlines from obligations imposed by law or an applicable agreement.
For software implementing an invention covered by this policy, please clarify which developments concerning publication or third-party use require continuing updates under page 3. Where that software is openly distributed, routine downstream reuse should not itself create an ongoing reporting burden unless relevant to a specifically identified obligation. This requirement should not apply to nonpatentable software excluded from the policy.
7. Preserve Student Protections and the Team's Accepted Research Plan
The examples on page 2 are helpful, but their exclusion of work that builds upon or improves existing University ITRM needs clarification. Please explain how this qualification applies to student work involving a covered invention, and confirm that improving university-developed software does not itself bring an otherwise excluded copyrightable contribution within this policy.
The current patent policy requires the PI’s written acceptance before GW enters a sponsored research agreement containing provisions that differ from university policy. The draft’s “Externally-Funded Research” section (pages 4–5) retains the PI’s duty to inform colleagues of their rights and duties, but does not expressly retain that acceptance requirement. Please preserve the requirement or identify where an equivalent binding safeguard is maintained. This matters where proposed sponsor terms would restrict software dissemination or alter the allocation of rights.
8. Clarify Databases, Continued Access, and Governance
Page 6 defines Tangible Research Materials as physical research products but includes computer databases. Please distinguish physical materials, digital data, copyrightable selection or arrangement, and database software. These involve different rights and obligations. Please explain how this definition operates alongside the copyright exclusion.
On pages 4 and 8, abandonment notice is welcome, but discretionary return following an inventor's request is not equivalent to the current required offer of assignment before abandoning a patent application. Please preserve an offer obligation, subject to applicable legal and sponsor constraints, with decision deadlines and written reasons. Open licensing or another access mechanism may sometimes better serve continued research than transfer of ownership.
The explicit research and teaching reservation on page 4 concerns returned inventions. Please clarify whether ordinary external licenses preserve the university’s and other nonprofit institutions’ research rights, as contemplated in the current policy’s non-federal sponsor provisions. For software-related inventions, we also recommend reserving rights sufficient for continued teaching and research software development.
Appeals remain under page 9, but the proposed panel replaces the current specified school and committee representation with EROC and at least one Faculty Senate Research Committee member. Dispute procedures also sit outside the policy itself. Please ensure meaningful faculty representation, access to software expertise, appropriate recusal, and shared governance over consequential changes to both policy and procedures. Essential scholarly rights should appear in the policy itself.
Corollary Recommendation: Establish Software Support and Coordination
Separately from the substantive policy protections above, GW should clarify the OSPO's advisory role in helping researchers choose licenses, contribute to existing projects, plan releases, and sustain research software communities. Researchers should have access to that expertise without acquiring an additional mandatory clearance step.
The OSPO, TCO, and sponsored projects staff should establish a clear coordination process for the specific cases that require institutional action. That process should identify responsibilities, authorization and signing authority, and turnaround times. The enduring policy protection is researcher direction of scholarship; the supporting office arrangement can be documented in implementation guidance.
There are useful examples to draw on without assuming uniform peer practice. Johns Hopkins provides a joint OSPO–Technology Ventures workflow for contributor license agreements. Syracuse established its OSPO through collaboration among research, libraries, and IT, with technology transfer support. Carnegie Mellon's IP policy expressly gives academic freedom priority over possible financial rewards, although its software ownership rules differ from the approach requested here. These examples support the feasibility of coordinated services and mission-centered policy; they are not claims that peers share a single ownership model.
In closing, please revise the draft to recognize research software as scholarship, preserve PI and research-team direction, separate copyright and patent rights consistently, and provide practical open-source pathways. Public benefit and academic freedom should govern the administration of these provisions, supported by clear definitions, bounded restrictions, and timely review.
References
- Patents and Tangible Research Materials Policy Draft
- Current GW Patents and Scholarly Work Policy
- GW Copyright Policy
- AAUP Statement on Intellectual Property (academic-policy principles supporting scholarly dissemination and academic freedom):
- Johns Hopkins OSPO contributor license agreement workflow
- Syracuse OSPO establishment and institutional collaboration
- Carnegie Mellon Intellectual Property Policy