Are routine administrative activities limiting your ability to spend time and energy on the casework that really matters? In this blog post, we discuss real-world applications of artificial intelligence (AI) tools that can help you address common challenges and free up your time for more complex resettlement tasks requiring human judgment. This is Part Two in a two-part blog series on AI (click here to see Part One).
Note: Evaluate each AI tool against your organization’s privacy policies and requirements before use. The tools mentioned below are not an exhaustive list and are not endorsed by Switchboard over similar tools available.
Is my task appropriate for AI?
As you explore new artificial intelligence (AI) tools, remember that not every challenge is appropriate for AI. AI can help with routine or administrative tasks, brainstorming, and data analysis, but some situations require human judgment.
Ask yourself:
- Is this a high-stakes decision affecting newcomer outcomes? Examples include placement, benefits eligibility, safety, and immigration advice.
- Does this require cultural responsiveness or trauma sensitivity? Examples include complex case management, crisis intervention, and trust-building.
- Could AI errors cause serious harm? Examples include legal advice, medical consent, safety planning, and other situations requiring human judgment.
- Is there a simpler solution? Could simpler tools suffice, such as a better intake form, an FAQ sheet, or a process improvement?
If you answer “yes” to any of these questions, AI may not be the right tool. But for tasks where you answer “no” to each of the questions above, let’s explore below which AI tools best fit your workflow, organization, and clients.
How should I prepare before using an AI tool?
Before adopting any AI tool, work through these steps:
Define the Problem
Identify the specific challenge AI will address and quantify it when possible (e.g., “staff spends five hours a week translating documents”). Define success upfront, including expectations for accuracy, completeness, and efficiency (e.g., “translations are understandable, capture all essential information, and cut staff translation time in half”).
Understand How the AI Tool Uses Data
Assume any information you enter into AI tools is not private unless you obtain written confirmation otherwise.
- Configure privacy settings: Reduce risk by disabling data sharing, conversation history, memories, logging, and model training features whenever possible. Note: Adjusting these settings will improve privacy but does not guarantee complete security.
- Verify organizational compliance: Before using any AI tool, confirm that it aligns with your organization’s privacy policy and any AI-specific policies your organization may have in place. For guidance, see Switchboard’s resource Using Artificial Intelligence in Service Delivery: A Framework to Evaluate Organizational Readiness.
Protect Client Privacy and Confidentiality
Only share the minimum information needed, and never enter personally identifiable or sensitive client data into a new AI tool. If you would like to use an AI tool in service delivery, explain its use in plain language to the client and obtain their consent whenever possible. For guidance, see Switchboard’s resource Talking with Newcomers about Artificial Intelligence (AI).
Keep a Human in the Loop
AI outputs should always be reviewed by staff. Use the SIFT Method, a quick, four-step check for evaluating AI-generated content:
- Stop before acting on AI-generated content
- Investigate the sources behind the information
- Find trusted sources to confirm key claims
- Trace information back to its original context
Document when staff review, verify, or override AI outputs, and never allow AI to make final decisions affecting clients.
Understand Data Privacy and Payment Tiers
Before entering client information into an AI tool, confirm which version of the tool your organization uses. Not all AI subscriptions provide the same privacy protections. Free tools often use your inputs for training, while higher-tier plans may offer stronger safeguards—but never assume that any tier is private. Always verify your organization’s plan and privacy settings before entering client-related information.
What are some specific AI tools to try?
The tools below are designed for broad commercial use and may not always align with the needs of resettlement services. Before adopting any tool, consider how well it fits your organization’s values, clients, and goals.
| Translation and Client Communication Tools | |
|---|---|
| Tarjimly – A nonprofit connecting refugees with volunteer interpreters, using AI-enhanced matching | DeepL – Offers text-based language translation and is often more accurate than Google Translate, especially for European languages |
|
Tip: Never use AI-only translation for legal documents, medical consent, or critical communications. Always have bilingual staff verify outputs and be aware that accuracy varies significantly by language.
|
|
| Brainstorming and Design Tools | |
|---|---|
| Claude, Copilot, Gemini, or ChatGPT – Generate activity ideas, explore different approaches, update communication to newcomers in plain language | NotebookLM – Generate audio podcasts, slides, quizzes, infographics based only on the information/sources that you provide |
|
Tip: AI outputs reflect the context you provide. Generic prompts tend to result in generic results, while more specific prompts produce more useful responses. Practice context engineering, which means giving the AI model the right context, in the right order, for the task and leaving out any information it should not have. When using AI for newcomer services, provide context about your audience, constraints, and goals. Instead of asking AI to “create a flyer,” ask it to improve or build on materials you already have.
|
|
| Note-Taking and Summarizing Tools | |
|---|---|
| Magic Notes – AI-powered meeting summarization tool that uses templates for documentation | Otter.ai – Transcribes conversations; creates searchable notes |
|
Tip: These tools require consent to record from anyone included in the meeting and have features only available in the enterprise versions (not free consumer tools). Don’t forget to have staff review all outputs for accuracy.
|
|
What’s next?
Ultimately, AI tools can help reduce administrative burden, support communication, and spark new ideas, but they are most effective when paired with strong human judgment, privacy protections, and thoughtful oversight. By starting with clearly defined problems, protecting client confidentiality, and keeping humans in the loop, resettlement staff can explore AI in ways that are practical, ethical, and responsive to newcomer needs.
Resources
- Toolkit: Using AI in Service Delivery – A Readiness Framework
- Archived Webinar: Moral Imagination: From Smart to Wise AI
- MicroLearning Video: Creating Quick Tutorial Videos for Clients
- Archived Webinar: Beyond the Code: Responsible Technology for Refugee Resilience
- Blog: Search Smarter: Finding Research Evidence through Responsible Use of AI
- Blog: Insights from the Field: The Use of Artificial Intelligence in Resettlement Work
- Archived Webinar: Building Self-Sufficiency with AI
- Blog: Innovative Ways Newcomers Are Using Technology
- Podcast: The Role of Tech in Resettlement: Discussing Artificial Intelligence
The IRC received competitive funding through the U.S. Department of Health and Human Services, Administration for Children and Families, Grant #90RB0053. The project is 100% financed by federal funds. The contents of this document are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services, Administration for Children and Families.
