Throughout the past 18 months, the Goldhirsh Foundation and its LA2050 initiative have provided complimentary AI consulting, training, and workshops to the Los Angeles impact community. The work is led on the Foundation’s behalf by Jen García, Goldhirsh Foundation’s AI executive-in-residence.

Below – and in upcoming monthly posts – García shares practical insights, real-world use cases, and emerging research to support nonprofits’ responsible and ethical adoption of AI tools.

Getting Started

First and foremost, in order for nonprofit leaders and staff to begin responsible, ethical AI adoption, they must remember that despite urging to jump in from individual AI champions and some media narratives, it is key to thoughtfully consider three key elements before diving in:

1. This is not a sprint; it’s a marathon. Yes, generative AI has been on the scene commercially for two and a half years, but that doesn’t mean nonprofit organizations should jump in head-first. We encourage leaders to take a deliberate, steady approach to adoption, especially due to the potential risks.
2. Ensure that leadership undergoes training first. It is vital that leadership have an overview of generative AI tools and learn basic prompting before they make decisions on when and how it can be used for work. This training will also inform writing a working AI policy. We will cover these issues more in-depth in upcoming blogs.
3. Start with low-risk, low-stakes tasks. Once leadership is trained on generative AI basics, they will be able to identify use cases that align with the nonprofit’s mission and values. For example, although generative AI tools are often competent at producing marketing copy, that is not a good use case for nonprofits, who have deep community connections and unique trusted voices to uphold.

Two Low Risk, Low Stakes Tasks

1. ChatGPT’s Projects are a useful way to not only group common threads but also upload files for continual context and provide custom instructions each prompt will adhere to within the project. This feature is available in paid subscriptions within ChatGPT Plus, Teams or Enterprise. (In fact, nonprofits receive 20% off of ChatGPT Teams.)
2. The Goldhirsh Foundation provided support to the Los Angeles Fire Department Foundation, who in turn introduced us to LAFD Captain Matthew Sanchez. He told us he found it useful to leverage Google’s NotebookLM to research – or just search – through publicly available information. (NotebookLM is a free generative AI tool that, unlike other tools, focuses solely on the context you provide.) Rather than searching through PDFs individually, Captain Sanchez can upload them into NotebookLM and query through them all at once. Nonprofit leaders may find this useful for grant or donor research, among many research-based use cases.

Photo: Participants at Hands-On AI Training for Nonprofits, a workshop held in March 2025 by the Goldhirsh Foundation and LA2050.