How AI is Changing Automation: What Nonprofit Leaders Need to Know
Table of Contents
AI is becoming deeply woven into our daily lives, influencing the way we learn and work. It’s taking traditional automation to the next level, enhancing output, optimizing workflows, and unlocking new insights across all industries.
There is broad optimism and consensus around the potential benefits of AI for productivity, but many nonprofits are concerned about data security and social impacts.
Seeing widespread interest in AI in the nonprofit sector, and building on our experience using it in accounting and finance functions, YPTC hosted a webinar covering:
- Evolution of Automation and AI
- Use Cases of AI for Nonprofits
- Use Cases of AI for Accountants
- Data Privacy, Compliance, and Risk Management
- Ethical Considerations
- Best Practices for Nonprofit Leaders
- How to Get Started
- Additional Resources
In this guide, we summarize key takeaways from the webinar and provide actionable insights that can help your organization navigate the AI transformation smoothly, wherever you are in your journey.
Evolution of Automation and AI
Automation has long helped nonprofits streamline repetitive tasks, like processing survey responses or generating basic reports from your information systems. But traditional automation relies on fixed rules—think “if this, then that”—and struggles with tasks that require judgment or context. Using Power Automate to pull data from multiple sources into a single workbook is an example of standard automation.
Machine learning levels up. Instead of just following rules, machine learning can make its own rules by learning patterns from examples. In a nonprofit organization, machine learning could be used to analyze patterns in survey response data and classify sentiment. In everyday life, Netflix recommendations, facial recognition for unlocking smartphones, and email spam filters are all examples of machine learning algorithms.
Generative AI goes beyond analysis. It processes natural language questions and generates human-like responses by training on massive, unlabeled data sets. For example, generative AI can recommend actions in response to survey data that are aligned with your organization’s mission, strategic plan, and board-approved budget. Large language models like ChatGPT are an example of generative AI.
A drawback of generative AI is that the tools are limited by what they’ve learned (their core knowledge), and, despite this limitation, when you give them a prompt, they have to respond. This means when they don’t have a reasonable response, they give you their best guess—often referred to as a hallucination.
Reasoning models can prompt and re-prompt themselves to check their responses internally, reduce errors and hallucinations, and deliver validated answers.
Agentic AI goes even further, connecting with other tools (like your accounting system or the web) to gather real-time information and automate complex workflows. Agentic AI represents a shift from AI as a single tool to AI as an orchestrator of multiple tools.
Now we can create AI agents that perform multiple complex reasoning steps so we don’t have to manually re-prompt to add more context and force the AI to check its reasoning. Copilot’s Researcher agent, which uses chain-of-thought reasoning to iteratively refine its answers, is an example of an agent built with a reasoning model.
These advances mean AI can now help nonprofits answer questions like “Who is our biggest funder this year?”—purely from a natural language prompt—by writing code to analyze data across multiple systems, executing calculations, and generating a clear, accurate, conversational response.
Let’s look at some practical applications of automation and AI for nonprofits.
Use Cases of AI for Nonprofits
The use cases of AI for nonprofits are growing exponentially. Here are just a few of the tasks for which we’ve seen nonprofits using AI:
- Data entry and analysis
- Grant research
- Grant writing
- Note-taking and meeting summaries
- Volunteer matching and scheduling
Use Cases of AI for Accountants
Accountants are also increasingly using AI for a variety of tasks to boost efficiency, reduce errors, and enable greater focus on strategic and advisory activities. Here are a few examples:
- Invoice and bill processing – extracting data, matching to purchase orders, and flagging anomalies
- Bank reconciliations – matching data and flagging mismatches
- Tax research and preparation – identifying deductions, performing calculations, and answering complex regulatory questions
- Fraud detection – detecting anomalies and suspicious patterns in real time
- Forecasting and financial planning – using predictive analytics to create dynamic forecasts
Data Privacy, Compliance, and Risk Management
A recent survey showed that while most nonprofits are exploring or implementing AI tools, less than 25% of them have a strategy and even fewer have an acceptable use policy. Automation and AI must be used responsibly, not just to ensure the integrity of the output, but also to safeguard confidential data and maintain compliance with applicable regulatory requirements.
Take the following steps for responsible use of AI tools:
Choose enterprise-level solutions integrated within your organization’s secure IT environment (such as Microsoft 365 or Google Workspace). Do not upload confidential documents or share sensitive information on public AI tools.
Configure in accordance with the compliance standards to which your organization is otherwise committed, such as SOC 2, HIPAA, and GDPR. Always review privacy settings of AI tools and opt out of unnecessary data sharing.
Implement strong policies governing the use of all technology, including AI. You may have a separate policy for AI, or you may extend your acceptable use policy for digital tools and software in general to include AI.
Consult your legal team before deploying new AI tools. AI-generated content may not be eligible for copyright protection unless there’s significant human involvement, so keep a record of your contributions and review contracts carefully.
Assess risk before and after deployment. Certain use cases are inherently riskier than others. Here are some potential AI use cases and their typical, relative risk levels:
- Low risk: Using AI as an assistant to an individual who reviews the output – for example, brainstorming, copy editing, writing Excel formulas, or conducting web research
- Medium risk: Using AI as an internal representative or technical specialist – for example, answering questions from staff about company policies or analyzing data in response to natural language questions
- High risk: Using AI independently to execute financial transactions or communicate with external stakeholders – for example, approving expenses, reviewing resumes or grant applications, providing professional advice, or writing and posting on social media
Ethical Considerations
AI models can perpetuate social and cultural biases present in their training data. To mitigate this, nonprofits should:
- Regularly review AI outputs for bias and impact
- Gather input from a range of perspectives before deploying AI-generated content
AI also has an environmental footprint, requiring significant computing power and natural resources, including:
- Rapid and dramatic impacts on electrical energy production and consumption for data centers
- Extraction, mining, or intensive recycling of precious and rare earth metals that are used to build the computer chips that run AI models
- Significant water use to cool data centers
As technology advances, models are becoming more efficient, but it’s important to use AI judiciously and be mindful of its broader impact.
Best Practices for Nonprofit Leaders
- Start with clear guidelines and policies. If you don’t already have an AI policy, begin with your organization’s acceptable use policy for digital tools. Consider covering these topics in your policy:
- Data privacy and security requirements
- Appropriate use cases for different AI tools
- Required human oversight and approval processes
- Documentation and auditing requirements
- Training and support resources
- Align AI use with your mission and values. Consider the social and environmental impacts, and ensure your team feels comfortable and empowered.
- Monitor and review. Regularly assess the impact of AI on your staff, clients, and stakeholders, and adjust your approach as needed.
Ready to Embrace AI? Here’s How to Get Started:
- Identify repetitive or judgment-based tasks that could benefit from automation or AI.
- Explore secure, enterprise-grade AI tools within your existing IT environment.
- Train your team on acceptable use of AI tools and prompting best practices.
Additional Resources
For a deeper dive on this topic, watch the webinar, How AI is Changing Automation.
Download YPTC’s Mindful Prompting Checklist Here
Have questions or want personalized guidance? Contact YPTC to learn how AI can support your nonprofit’s mission.
Fill out this form to download the Mindful Prompting Checklist:
