{"id":657218,"date":"2025-03-15T15:10:14","date_gmt":"2025-03-15T12:10:14","guid":{"rendered":"https:\/\/en.buradabiliyorum.com\/open-model-licenses-often-carry-concerning-restrictions\/"},"modified":"2025-03-15T15:10:14","modified_gmt":"2025-03-15T12:10:14","slug":"open-model-licenses-often-carry-concerning-restrictions","status":"publish","type":"post","link":"https:\/\/buradabiliyorum.com\/en\/open-model-licenses-often-carry-concerning-restrictions\/","title":{"rendered":"#’Open’ model licenses often carry concerning restrictions"},"content":{"rendered":"
\n

This week, Google released a family of open AI models, Gemma 3, that quickly garnered praise for their impressive efficiency. But as a number<\/a> of<\/a> developers<\/a> lamented on X, Gemma 3\u2019s license makes commercial use of the models a risky proposition.<\/p>\n

It\u2019s not a problem unique to Gemma 3. Companies like Meta also app<\/a>ly custom, non-standard licensing terms to their openly available models, and the terms present legal challenges for companies. Some firms, especially smaller operations, worry that Google and others could \u201cpull the rug\u201d on their business by asserting the more onerous clauses.<\/p>\n

\u201cThe restrictive and inconsistent licensing of so-called \u2018open\u2019 AI models is creating significant uncertainty, particularly for commercial adoption,\u201d Nick Vidal, head of community at the Open Source Initiative, a\u00a0long-running institution<\/a>\u00a0aiming to define and \u201csteward\u201d all things open source, told TechCrunch. \u201cWhile these models are marketed as open, the actual terms impose various legal and practical hurdles that deter businesses from integrating them into their products or services.\u201d<\/p>\n

Open model developers have their reasons for releasing models under proprietary licenses as opposed to industry-standard options like Apache and MIT<\/a>. AI startup Cohere, for example, has been clear<\/a> about its intent to support scientific \u2014 but not commercial \u2014 work on top of its models.<\/p>\n

But Gemma and Meta\u2019s Llama\u00a0licenses\u00a0in particular have restrictions\u00a0that limit the ways companies can use the models without fear of legal reprisal.<\/p>\n

Meta, for instance, prohibits developers<\/a> from using the \u201coutput or results\u201d of Llama 3 models to improve any model besides Llama 3 or \u201cderivative works.\u201d It also prevents companies with over 700 million monthly active users from deploying Llama models without first obtaining a special, additional license.<\/p>\n

Gemma\u2019s license<\/a> is general<\/a>ly less burdensome. But it does grant Google the right to \u201crestrict (remotely or otherwise) usage\u201d of Gemma that Google believes is in violation of the company\u2019s prohibited use policy<\/a> or \u201capplicable laws and regulations.\u201d\u00a0<\/p>\n

These terms don\u2019t just apply to the original Llama and Gemma models. Models based on Llama or Gemma must also adhere to the Llama and Gemma licenses, respectively. In Gemma\u2019s case, that includes models trained on synthetic data generated by Gemma.<\/p>\n

Florian Brand, a research assistant at the German Research Center for Artificial Intelligence, believes that \u2014 despite what tech giant execs would have you believe<\/a> \u2014 licenses like Gemma and Llama\u2019s \u201ccannot reasonably be called \u2018open source.\u2019\u201d<\/p>\n

\u201cMost companies have a set of approved licenses, such as Apache 2.0, so any custom\u00a0license\u00a0is a lot of trouble and money,\u201d Brand told TechCrunch. \u201cSmall companies without legal teams or money for lawyers will stick to models with standard licenses.\u201d <\/p>\n

Brand noted that AI model developers with custom licenses, like Google, haven\u2019t aggressively enforced their terms yet. However, the threat is often enough to deter adoption, he added.<\/p>\n

\u201cThese restrictions have an impact on the AI ecosystem \u2014 even on AI researchers like me,\u201d said Brand.<\/p>\n

Han-Chung Lee, director of machine learning at Moody\u2019s, agrees that custom licenses such as those attached to Gemma and Llama make the models \u201cnot usable\u201d in many commercial scenarios. So does Eric Tramel, a staff applied scientist at AI startup Gretel.<\/p>\n

\u201cModel-specific licenses make specific carve-outs for model derivatives and distillation, which causes concern about clawbacks,\u201d Tramel said. \u201cImagine a business that is specifically producing model fine-tunes for their customers. What\u00a0license\u00a0should a Gemma-data fine-tune of Llama have? What would the impact be for all of\u00a0their\u00a0downstream customers?\u201d<\/p>\n

The scenario that deployers most fear, Tramel said, is that the models are a trojan horse of sorts.<\/p>\n

\u201cA model foundry can put out [open] models, wait to see what business cases develop using those models, and then strong-arm their way into successful verticals by either extortion or lawfare,\u201d he said. \u201cFor example, Gemma 3, by all appearances, seems like a solid release \u2014 and one that could have a broad impact. But the market can\u2019t adopt it because of its\u00a0license\u00a0structure. So, businesses will likely stick with perhaps weaker and less reliable Apache 2.0 models.\u201d<\/p>\n

To be clear, certain models have achieved widespread distribution in spite of their restrictive licenses. Llama, for example, has been downloaded hundreds of millions of times and built into products from major corporations, including Spotify.<\/p>\n

But they could be even more successful if they were permissively licensed, according to Yacine Jernite, head of machine learning and society at AI startup Hugging Face. Jernite called on providers like Google to move to open license frameworks and \u201ccollaborate more directly\u201d with users on broadly accepted terms.<\/p>\n

\u201cGiven the lack of consensus on these terms and the fact that many of the underlying assumptions haven\u2019t yet been tested in courts, it all serves primarily as a declaration of intent from those actors,\u201d Jernite said. \u201c[But if certain clauses] are interpreted too broadly, a lot of good work will find itself on uncertain legal ground, which is particularly scary for organizations building successful commercial products.\u201d<\/p>\n

Vidal said that there\u2019s an urgent need for AI models companies that can freely integrate, modify, and share without fearing sudden\u00a0license\u00a0changes or legal ambiguity.<\/p>\n

\u201cThe current landscape of AI model licensing is riddled with confusion, restrictive terms, and misleading claims of openness,\u201d Vidal said. \u201cInstead of redefining \u2018open\u2019 to suit corporate interests, the AI industry should align with established open source principles to create a truly open ecosystem.\u201d<\/p>\n<\/div>\n

If you liked the article, do not forget to share it with your friends. Follow us on\u00a0Google News<\/a><\/span>\u00a0too, click on the star and choose us from your favorites.<\/span><\/strong><\/p><\/blockquote>\n

\n

If you want to read more like this article, you can visit our Technology<\/a><\/span> category.<\/strong><\/p>\n<\/blockquote>\n

Source<\/a><\/span><\/p>\n","protected":false},"excerpt":{"rendered":"

This week, Google released a family of open AI models, Gemma 3, that quickly garnered praise for their impressive efficiency. But as a number of developers lamented on X, Gemma 3\u2019s license makes commercial use of the models a risky proposition. It\u2019s not a problem unique to Gemma 3. Companies like Meta also apply custom,…<\/p>\n","protected":false},"author":1,"featured_media":657219,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"fifu_image_url":"https:\/\/techcrunch.com\/wp-content\/uploads\/2024\/01\/GettyImages-1548038240.jpg?resize=1200,814","fifu_image_alt":"","footnotes":""},"categories":[18],"tags":[77337,74329,153946],"class_list":["post-657218","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-technology","tag-ai","tag-licensing","tag-open-models"],"_links":{"self":[{"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/posts\/657218","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/comments?post=657218"}],"version-history":[{"count":0,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/posts\/657218\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/media\/657219"}],"wp:attachment":[{"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/media?parent=657218"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/categories?post=657218"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/tags?post=657218"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}