Meta has officially announced the recipients of its 2025 Llama Impact Grant, demonstrating its commitment to advancing the responsible and ethical use of artificial intelligence (AI) for social good. The Llama Impact Grant aims to support organisations and research institutions that are using AI to tackle significant challenges in education, health, environmental sustainability, and equitable access to technology. This initiative not only offers financial support but also facilitates collaborative engagement with Meta’s technical and research teams, enabling grantees to push the boundaries of innovation. The 2025 programme has gained attention for its focus on projects that address pressing global issues, prioritising underrepresented regions and communities.

The Llama Impact Grant is named after Meta’s open-source large language model family, LLaMA, which lies at the heart of this initiative. Meta made the LLaMA model publicly available to empower external developers, academics, and non-profits to innovate freely. This open-source philosophy is a key aspect of Meta’s approach, and it is reflected in the grant, which supports projects using LLaMA tools to create practical, impactful solutions. The programme not only provides resources but also encourages the development of AI applications that are ethical and community-driven, helping to shape the future of technology in a way that benefits society as a whole.

In 2025, 15 organisations across five continents received the Llama Impact Grant, each undertaking projects as diverse as the regions they represent. Some notable recipients include the Alan Turing Institute in the UK, which will use the funding to develop equitable AI frameworks for public policy decision-making. In Africa, Masakhane is using the grant to expand its research into underrepresented languages, enabling AI tools to become more accessible to millions of people. In Brazil, the Instituto de Tecnologia e Sociedade is building an AI-powered educational toolkit for teachers in low-income schools, allowing them to integrate AI literacy into their classrooms. Another recipient, DataKind, is applying AI to develop early warning systems for natural disasters in Bangladesh and Indonesia, while the AI for Good Foundation in the USA is creating an AI ethics curriculum to educate students about the real-world implications of AI development.

Despite the varied nature of these projects, certain common themes emerge. A clear focus on equity and inclusion runs through many of the initiatives, with several projects targeting underserved communities or underrepresented groups in the tech space. Education and literacy also take centre stage, with multiple projects designed to improve access to AI tools and foster a better understanding of these technologies. Transparency and accountability in AI development are other recurring themes, with grantees prioritising ethical AI practices to ensure that their tools are both fair and responsible. Several recipients are also addressing environmental sustainability, reflecting a growing concern for the planet’s future and the role AI can play in mitigating climate change and monitoring ecological health.

The Llama Impact Grant represents more than just a philanthropic gesture; it is part of a broader strategic move by Meta to position itself as a leader in responsible AI development. For businesses, developers, and organisations focused on AI or social impact, the announcement highlights the increasing importance of ethical development and collaboration. The Llama models, now at their third iteration, have been designed to be more accessible to a wider range of users, including those with limited resources. The open-source nature of LLaMA models ensures that even smaller organisations, constrained by infrastructure limitations, can build and deploy AI tools without needing the vast computational resources typically required for such projects.

As Meta continues to invest in responsible AI and open-source tools, the company’s approach is positioning itself to distinguish itself from other tech giants like OpenAI and Google. While these companies focus on commercialising their AI models, Meta is focusing on creating a more inclusive and transparent ecosystem that encourages collaboration for the public good. This focus on ethical AI is not without its challenges—questions regarding data sourcing, model bias, and the risks associated with open access remain—but Meta’s commitment to addressing these issues is clear. The company’s documentation around Llama 3 has improved, and it is working with ethical oversight bodies to ensure its models are used responsibly.

Looking to the future, the Llama Impact Grant Recipients is expected to evolve further. Although applications for the 2025 cohort have closed, Meta is likely to renew the programme in 2026, providing additional opportunities for organisations to secure funding and support. Those interested in applying for future rounds should focus on projects with clear, measurable impacts, a commitment to ethical AI practices, and an emphasis on community-driven solutions. The Llama models themselves are freely available for use, meaning that developers and organisations can start experimenting and contributing to the growing body of open-source AI knowledge today.

The 2025 Llama Impact Grant recipients signify a shift in how AI tools are applied and funded. Meta’s emphasis on equity, education, and transparency has set a new benchmark for how large technology companies can engage with the global community to drive positive change. As AI continues to evolve and influence various aspects of life, initiatives like the Llama Impact Grant offer a glimpse into a future where technology is used not only for commercial gain but for the greater good of society. By supporting mission-driven projects, Meta is helping to ensure that the AI tools of tomorrow are developed with fairness, responsibility, and accessibility at their core.