[License-review] For Approval: OpenMDW License Agreement, versions 1.1 (OpenMDW-1.1)

[Moming Duan]

McCoy,

Let me try to answer your second question. The difference I see is not about litigation in the abstract. It is about who can know what, before the lawsuit happens.

A termination clause works as a deterrent only when the party giving up its claims can know in advance what it is giving up. That is roughly true in the patent cases: an armed party can study the landscape and decide whether to sue and lose the license, or keep using the work.

With a model, the party giving up claims is the user, and the user cannot know. Here is the technical background. In a traditional software copyright dispute, the holder can read the source code, run a code review, and point at the copied lines. A model offers nothing to read: the weights are huge tables of numbers. And in my RAG example, my text has been converted into vectors, so the database holds neither my sentences nor anything a reviewer could match against my work. Without using the model, I have no way to know whether my work ended up in there at all.

My earlier example makes this concrete: I use a model through an API and find that my inputs are being stored in the provider's database for RAG. Under the license definition, that database would likely count as associated data, part of the Model Materials. So when I claim the copyright of my own work stored there, I am claiming that the Model Materials infringe my copyright, and my license terminates. I could not have known this before accepting the license, and I only find out while using the model.

The warranty disclaimer does not cover this. A disclaimer leaves me to bear risks from third parties, and I keep my claims against anyone. Paragraph 5 ends my license exactly when I bring a claim against the licensor itself. A risk I must bear is not the same as a claim I am punished for making.

So from where I stand, this clause does not work like the patent clauses it borrows from. It deters an ordinary copyright holder who had no way to check first, on a work he cannot inspect.


Best,
Moming Duan


From: License-review <license-review-bounces at lists.opensource.org> on behalf of McCoy Smith <mccoy at lexpan.law>
Date: Monday, August 31, 2026 at 08:04
To: license-review at lists.opensource.org <license-review at lists.opensource.org>
Subject: Re: [License-review] For Approval: OpenMDW License Agreement, versions 1.1 (OpenMDW-1.1)



On 8/30/2026 2:50 PM, Pamela Chestek wrote:
McCoy, you asked for a definition of "software freedom" and Moming's response is similar to my thoughts:

On 8/29/2026 11:10 PM, Moming Duan wrote:

Here is a simple test of my own: if 99% of the world's models were released under license X, would our ecosystem be better off?

My premise is that the everyone benefits from a software commons, so, does the license encourage or discourage sharing? "Software freedom" is the ability to do anything you want with the software. A few conditions/impairments are allowed because they indirectly encourage the growth of the software commons - giving credit (copyright notice/attribution), telling people what their rights are (the license), the appropriate distribution of risk (the "as-is" rep and disclaimer of warranty), and compulsory sharing (copyleft). The balance is very delicate. It's questionable whether the GPLv3, and then the AGPL, contribute to the health of the open source software commons because they are commonly used as the antagonist to encourage people to get a commercial license instead. It does, though, seem that lever isn't as successful anymore, since people are more willing to comply with the copyleft licenses and these two licenses are more often being used for their intended benefit.

This, I think, is the piece that's missing from the OSD. We've seen the OSD gamed when it has been formally met but it is nevertheless clear that the license is designed to give one entity superior rights in the software or has mechanisms that will discourage the use of the software. "Providing software freedom" as a requirement is a check on that.

I guess the two issues I have with this are:

1. How does one who is trying to write a license ever know what to say or not say in their license to satisfy this test? At least the OSD (and to a certain extent, the FSD) articulate tests against which one can measure the text and the rights and obligations of a particular license. There are plenty of people who believe that non-copyleft makes the ecosystem better off, and plenty of others that think copyleft does that. I think end the end, OSI has arrived at the position, via the OSD, that neither are correct and both models make the ecosystem better off.

2. Specifically, the primary objection to OpenMDW-1.1 seems at this point to be that it allows termination of the license upon assertion of a copyright infringement claim against the licensed work. We already know, I think at this point, that terminating the patent license part of a license is an appropriate termination event upon a patent assertion (lots of OSI-approved licenses, including popular ones, do this), and in at least two examples (GPLv3 & MPLv2), terminating the copyright license part of the license is an appropriate termination event upon a patent assertion. Why isn't *any* litigation against something licensed open source (except, I suppose, litigation to ensure compliance with the license itself) something that impinges upon "the ability to do anything you want with the software"? Getting sued (or even the threat of being sued) is an impediment to doing what you want with a licensed work. Whether it be for patent infringement, for personal injury, for breach of warranty, or even copyright infringement. In fact, many open source licenses disclaim warranties of non-infringement. Meaning, you can't sue the licensor for a work that you received under an open source license that results in a third party asserting an IP claim against the work. Why isn't that warranty disclaimer equally problematic?

To my mind, the carve out that is being argued here -- it is only copyright claims against the licensor by the licensee that it is impermissible to have the license try to prevent -- seems solely driven by the perception that: a) AI models are rife with copyright infringement; b) copyright infringement claims against those models are inherently meritorious, or at least, ought not be impeded; c) the calculus that many, many open source licenses make re patents -- you can sue, or you can continue to use the license, but not both -- shouldn't equally apply to copyrights; d) AI model creators are trying to "game the system" in a way that is somehow more of a risk than the "system gamers" of the past (for example, the antagonist model you describe above).

At a minimum, it'd be useful to get ourselves down to a point where we articulate *all* the conditions in this particular license that are perceived to be anti-software freedom. Seems like the termination upon copyright assertion is one that seems to be the main focus at this point, but I'd like to at least understand if there are any others (such as the definition of the licensed works, or the "you are solely responsible for" clause). It seems at this point those might have fallen away as concerns, but I want to make sure so that when/if this comes up for a Board decision, all the relevant concerns are discussed.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensource.org/pipermail/license-review_lists.opensource.org/attachments/20260831/2d5dba70/attachment-0001.htm>