《Linux基金会：2025全球开源AI合作战略方向研究报告（英文版）（31页）.pdf》由会员分享，可在线阅读，更多相关《Linux基金会：2025全球开源AI合作战略方向研究报告（英文版）（31页）.pdf（31页珍藏版）》请在三个皮匠报告上搜索。

1、Cailean Osborne,PhD,The Linux FoundationForeword by Nicolas Flores-Herr,PhD,Fraunhofer IAISJuly 2025Charting Strategic Directions for Global Collaboration in Open Source AIKey Takeaways from the GOSIM Open Source AI Strategy ForumSponsored byOpen source developers influence responsible AI practices

2、by sharing tools and documentation that others can use,learn from,and build on.Developers should champion open development practices and licenses that uphold the four freedoms of open source in open source AI.Governments should embrace open source AI as a strategic tool to foster both global collabo

3、ration and competition in AI.Small models increasingly rival the performance of large models while providing affordability,customizability,and privacy benefits.The open science approach to AI R&D facilitates reproducibility,knowledge sharing,learning,and auditing of potential biases and security vul

4、nerabilities.The success of open source AI relies not only on sharing models,code,and datasets but also on building and sustaining a vibrant global community.Developing open datasets for medium-and low-resource languages is a key priority for tackling cultural and linguistic biases in LLMs.Open sour

5、ce evaluation frameworks and benchmarks for diverse capabilities and contexts are key to enhancing trust and safety in open source AI.Access to public compute infrastructure is a major bottleneck for the participation of researchers and grassroots initiatives in open source AI.Innovative funding mec

6、hanisms and business models are needed to support the growth and sustainability of open source AI startups.Neutral foundations can foster collaboration between global enterprises on open source tools and open standards across the AI stack.There is a narrow window of opportunity to develop interopera