If you want to add restrictions to the MIT license, just don't use the MIT license or name it that way. Your change makes it no longer an open or permissive license so it's bizarre to keep the name. It would be like creating the MIT But-You-Have-To-Pay License.
Then lets call this special license something like:
No-you-still-dont-have-to-pay-but-any-AI-use-is-restricted-license.
1
This way, everyone knows, its not as free as the MIT license and has absolutely no relation to it. /s
Ofcourse you can specialize existing licenses with limited paragraphs and reflect that in their names...
I think you're missing the point of the MIT License and its history. If you want a proprietary license, "specializing" the MIT License for that is silly. My grandmother could specialize as a bicycle if I added wheels to her.
Yes, the MIT ought to be as broad, free and compatible as possible and is imo targeted to counter copy left licenses but this permissiveness can be also a problem, so specializing it into something like a tagged creative commons license can be a reasonable effort. It depends on the problem/topic of restriction and how narrow or well defined they are.
> Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software
Key words are:
- permission is [...] granted
- free of charge
- without restriction
- use, copy, …
Then:
> may not be used for the purposes of […]
The license contradicts itself.
> Don't we have to ask for permission before feeding someone's years of work into an AI?
That's the point of an OpenSource license, to give permission.
This kind of stuff makes me think very few people really understand what OpenSource is about. The very same people who will fallback to licenses such as the BSL as soon as people/companies will use the permissions that they gave, and then will complain that "no one wants to pay for the thing i did for free and nobody asked for".
The AI Act will be fully applicable from 2 August 2026.
Providers of GPAI models must respect Text and Data Mining (TDM) opt-outs.
2.1 Legal Basis: Article 53(1)(c) AI Act and Directive (EU) 2019/790
The Copyright Chapter of the Code directly addresses one of the most contentious legal questions in AI governance: the use of copyrighted material in training GPAI models and the risk of infringing outputs. Article 53(1)(c) AI Act requires GPAI providers to “identify and respect copyright protection and rights reservations” within their datasets.
+
This obligation complements the framework of Directive (EU) 2019/790 on copyright and related rights in the Digital Single Market (DSM Directive). Notably, Article 4(3) DSM Directive allows rightsholders to exclude their works from text and data mining (TDM) operations via machine-readable opt-outs.
Well, in the eyes of the EU, the entire 'fair use' thing doesn't exist there at all (per the EU's own JURI, which I think is roughly the equivalent of the U.S.'s Office of the Attorney General, with similar duties around defining the canonical interpretations of the laws).
Two relevant bits, dug out from the 175-page whole:
> Although the Act tries to
address this by extending obligations to any provider placing a GPAI model on the EU market, the
extraterritorial enforcement of these obligations remains highly uncertain due to the territorial nature
of copyright law and the practical difficulty of pursuing infringement claims when training occurs under
foreign legal standards, such as U.S. fair use.
and:
> Finally, it is important to clarify that the current EU framework provides a closed list of exceptions and
does not recognise a general fair use defence. As a result, AI-generated outputs that include protected
expression without a valid exception remain unlawful.
It seems to be dated the same month as DDG's analysis, July 2025, so I would expect the MIT Non-AI License that we're discussing here to be much more defensible in the EU than in the U.S. — as long as one studies that full 175-page "Generative AI and Copyright" analysis and ensures that it addresses the salient points necessary to apply and enforce in EU copyright terms. (Queued for my someday-future :)
From a U.S. standpoint: Licensing is a function of copyright. A work not subject to copyright cannot be licensed productively as-is, as the public domain quality of the work is a trivial and conclusive defense against a licensor’s claims of copyright violation. Fair use is not subject to copyright. Since licensing enforcement is only possible with an applicable copyright, enforcement cannot be completed against fair uses, as copyright law is not applicable to fair uses and therefore licensing enforcement has no legal basis. However, a judgment may overturn a defense of fair use brought against a defendant in a licensing enforcement claim, which would then subject the defendant’s use to copyright law and thus to license enforcement.
It’s midnight now, so you’re on your own to dig up and review specific instances of relevant case law, or to contrast with non-U.S. laws. Licensing above refers to i.e. LICENSE files of the specific sort that this post is about ("MIT Non-AI License"); other definitions of licensing, as well as e.g. DMCA exceptions, exist that might be of interest for you to explore further. I believe there’s been a handful of cases related to AI and fair use this past year, but as with all such defenses, unique circumstances are common enough that I hesitate to suggest any future outcome as 100% certain without much more case law than AI has today. (I am not your lawyer, this is not legal advice.)
Interesting perspective. I think you have a right, whats yours is yours to do as you please with - you know?
But can I give you another "viewpoint"? I guess it's like, "Wow, my code, my work, what came from my brain, my fingers" - it essentially lives forever, if you think about it - it becomes embedded and compressed inside weights/tokens. Like - part of you is there.
I guess it's cool. For me it's just to know that like this super intelligent things deep down actually knows who I am - my code is in it's architecture, it gives me feeling of honor in some way. You know?
- VC funded SF startups with an ethos of "Do whatever is necessary at first, scale then lawyer up once a unicorn and shits starts to hit the fan" this might try to prevent... and who will blissfully ignore it
- actually mindful actors, e.g public researchers, non-profit, etc who genuinely understand the point and will respect it but whom, ironically enough, you might want to support by providing them with more resources including your code
So... yes the intent is good but honestly I wouldn't trust the 1st category of actors to respect that. Technically what they "should" do, because it's safe and efficient, is only crawl the most popular compatible licenses, e.g. MIT, and ignore the rest. That's safe and pragmatic. Again what I expect them to do (just my hunch) is take EVERYTHING, get called on, apologize, change nothing, get brought to court, apologize again and only do something if the financial repercussion is bigger than the expected alternative. So probably still do nothing.
This should be about get permission from open-source developers before feeding their years of work into AI. I think we should not believe what Anthropic, OpenAI, Meta, Google tells.
Doesn't actually count as open source. Per the OSD, you cannot restrict the purpose for which people use the software and still be open source.
If you want to release your code as actual open source but legally restrict what AI companies do with it, use a copyleft license like the GPL. They could still use it for training, but the product of such training may itself fall under the GPL being a derivative work, and the AI corps don't want to touch that legal quagmire. (The GPL continues to be Stallman's brilliant legal aikido at work. Stallman is like that one guy from Final Fantasy Tactics Advance who gives you the power to manipulate the laws to work to your advantage.)
Honestly, it may be time to abandon open source as a concept. There are source-available strategies that cause less social harm than open source, including Coraline Ada Ehmke's "ethical source" and the Sentry project's "fair source" (https://fair.io) models that we can draw inspiration from.
If you want to add restrictions to the MIT license, just don't use the MIT license or name it that way. Your change makes it no longer an open or permissive license so it's bizarre to keep the name. It would be like creating the MIT But-You-Have-To-Pay License.
Then lets call this special license something like:
No-you-still-dont-have-to-pay-but-any-AI-use-is-restricted-license. 1 This way, everyone knows, its not as free as the MIT license and has absolutely no relation to it. /s
Ofcourse you can specialize existing licenses with limited paragraphs and reflect that in their names...
I think you're missing the point of the MIT License and its history. If you want a proprietary license, "specializing" the MIT License for that is silly. My grandmother could specialize as a bicycle if I added wheels to her.
Yes, the MIT ought to be as broad, free and compatible as possible and is imo targeted to counter copy left licenses but this permissiveness can be also a problem, so specializing it into something like a tagged creative commons license can be a reasonable effort. It depends on the problem/topic of restriction and how narrow or well defined they are.
This is what any AI agent tells
Yes, I'm a secret AI agent here to try and stop your powerful new licensing idea. /s
I meant I asked Gemini before and it told the same
> Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software
Key words are:
Then:> may not be used for the purposes of […]
The license contradicts itself.
> Don't we have to ask for permission before feeding someone's years of work into an AI?
That's the point of an OpenSource license, to give permission.
This kind of stuff makes me think very few people really understand what OpenSource is about. The very same people who will fallback to licenses such as the BSL as soon as people/companies will use the permissions that they gave, and then will complain that "no one wants to pay for the thing i did for free and nobody asked for".
Licenses have no bearing on fair use or where otherwise permitted to be ignored by law.
The AI Act will be fully applicable from 2 August 2026.
Providers of GPAI models must respect Text and Data Mining (TDM) opt-outs.
2.1 Legal Basis: Article 53(1)(c) AI Act and Directive (EU) 2019/790 The Copyright Chapter of the Code directly addresses one of the most contentious legal questions in AI governance: the use of copyrighted material in training GPAI models and the risk of infringing outputs. Article 53(1)(c) AI Act requires GPAI providers to “identify and respect copyright protection and rights reservations” within their datasets. + This obligation complements the framework of Directive (EU) 2019/790 on copyright and related rights in the Digital Single Market (DSM Directive). Notably, Article 4(3) DSM Directive allows rightsholders to exclude their works from text and data mining (TDM) operations via machine-readable opt-outs.
https://www.ddg.fr/actualite/the-european-unions-code-of-pra...
Well, in the eyes of the EU, the entire 'fair use' thing doesn't exist there at all (per the EU's own JURI, which I think is roughly the equivalent of the U.S.'s Office of the Attorney General, with similar duties around defining the canonical interpretations of the laws).
https://www.europarl.europa.eu/RegData/etudes/STUD/2025/7740...
Two relevant bits, dug out from the 175-page whole:
> Although the Act tries to address this by extending obligations to any provider placing a GPAI model on the EU market, the extraterritorial enforcement of these obligations remains highly uncertain due to the territorial nature of copyright law and the practical difficulty of pursuing infringement claims when training occurs under foreign legal standards, such as U.S. fair use.
and:
> Finally, it is important to clarify that the current EU framework provides a closed list of exceptions and does not recognise a general fair use defence. As a result, AI-generated outputs that include protected expression without a valid exception remain unlawful.
It seems to be dated the same month as DDG's analysis, July 2025, so I would expect the MIT Non-AI License that we're discussing here to be much more defensible in the EU than in the U.S. — as long as one studies that full 175-page "Generative AI and Copyright" analysis and ensures that it addresses the salient points necessary to apply and enforce in EU copyright terms. (Queued for my someday-future :)
Reference? A lot has changed within the last couple years.
From a U.S. standpoint: Licensing is a function of copyright. A work not subject to copyright cannot be licensed productively as-is, as the public domain quality of the work is a trivial and conclusive defense against a licensor’s claims of copyright violation. Fair use is not subject to copyright. Since licensing enforcement is only possible with an applicable copyright, enforcement cannot be completed against fair uses, as copyright law is not applicable to fair uses and therefore licensing enforcement has no legal basis. However, a judgment may overturn a defense of fair use brought against a defendant in a licensing enforcement claim, which would then subject the defendant’s use to copyright law and thus to license enforcement.
It’s midnight now, so you’re on your own to dig up and review specific instances of relevant case law, or to contrast with non-U.S. laws. Licensing above refers to i.e. LICENSE files of the specific sort that this post is about ("MIT Non-AI License"); other definitions of licensing, as well as e.g. DMCA exceptions, exist that might be of interest for you to explore further. I believe there’s been a handful of cases related to AI and fair use this past year, but as with all such defenses, unique circumstances are common enough that I hesitate to suggest any future outcome as 100% certain without much more case law than AI has today. (I am not your lawyer, this is not legal advice.)
Well the thing is, you have a copyright that you can license. However from what it currently looks like, fine-tuning/training is not copying.
Interesting perspective. I think you have a right, whats yours is yours to do as you please with - you know?
But can I give you another "viewpoint"? I guess it's like, "Wow, my code, my work, what came from my brain, my fingers" - it essentially lives forever, if you think about it - it becomes embedded and compressed inside weights/tokens. Like - part of you is there.
I guess it's cool. For me it's just to know that like this super intelligent things deep down actually knows who I am - my code is in it's architecture, it gives me feeling of honor in some way. You know?
Just my take.
I don’t always say AI pilled but
You deserve to be recognized beyond the false religion of the singularity.
Hilarious to get downvoted for defending the concept of
Literature
Without references and citations—- never mind.
Go read Animal Farm
I applaud the effort but honestly I see 2 cases :
- VC funded SF startups with an ethos of "Do whatever is necessary at first, scale then lawyer up once a unicorn and shits starts to hit the fan" this might try to prevent... and who will blissfully ignore it
- actually mindful actors, e.g public researchers, non-profit, etc who genuinely understand the point and will respect it but whom, ironically enough, you might want to support by providing them with more resources including your code
So... yes the intent is good but honestly I wouldn't trust the 1st category of actors to respect that. Technically what they "should" do, because it's safe and efficient, is only crawl the most popular compatible licenses, e.g. MIT, and ignore the rest. That's safe and pragmatic. Again what I expect them to do (just my hunch) is take EVERYTHING, get called on, apologize, change nothing, get brought to court, apologize again and only do something if the financial repercussion is bigger than the expected alternative. So probably still do nothing.
What if instead of a license most won’t respect, you include a poison pill in the repo or other code storage to poison the model?
https://news.ycombinator.com/item?id=45529587
https://news.ycombinator.com/item?id=45533842
This should be about get permission from open-source developers before feeding their years of work into AI. I think we should not believe what Anthropic, OpenAI, Meta, Google tells.
We should move to local LLMs.
How do local LLMs help?
Very ironically, this violates the MIT trademark in a quixotic pursuit of intellectual property ludditry.
For "AI", does that include some of the more advanced search indexing and auto complete tools?
Related, would you be in violation if you hosted this in a public GitHub repo, since it's in the TOS that they use source for training AI?
How do you propose someone who uses this license to enforce such clauses?
Article 4(3) of the Digital Single Market (DSM) Directive establishes a legal bridge between AI governance and copyright law.
https://www.ddg.fr/actualite/the-european-unions-code-of-pra...
Doesn't actually count as open source. Per the OSD, you cannot restrict the purpose for which people use the software and still be open source.
If you want to release your code as actual open source but legally restrict what AI companies do with it, use a copyleft license like the GPL. They could still use it for training, but the product of such training may itself fall under the GPL being a derivative work, and the AI corps don't want to touch that legal quagmire. (The GPL continues to be Stallman's brilliant legal aikido at work. Stallman is like that one guy from Final Fantasy Tactics Advance who gives you the power to manipulate the laws to work to your advantage.)
Honestly, it may be time to abandon open source as a concept. There are source-available strategies that cause less social harm than open source, including Coraline Ada Ehmke's "ethical source" and the Sentry project's "fair source" (https://fair.io) models that we can draw inspiration from.