AI and nonprofit jobs: How to review tasks, skills and workforce changes

AI and nonprofit jobs: How to review tasks, skills and workforce changes

AI and nonprofit jobs: How to review tasks, skills and workforce changes 150 150 Foundation List Nonprofit Jobs - Nonprofit, Foundation, Healthcare & Education Job Board

How should nonprofit teams respond to AI-related changes in work? Examine tasks, test specific uses, involve the people doing the work, and make staffing decisions from evidence. A model demonstration or economic forecast does not establish what will happen to a particular job, department, or community.

AI discussions often jump from a new capability to a prediction about employment. The missing step is how an organization chooses to use that capability. Systems, data, resources, service responsibilities, management, and review all shape what a workflow can do in practice.

This guide focuses on a practical workforce review. It helps leaders and employees discuss changes without assuming that every task will be automated or that technology adoption will produce a particular organizational result.

Separate a task from an entire job

A job usually combines activities with different requirements. A program coordinator might schedule meetings, maintain records, answer questions, coordinate partners, and make context-dependent judgments. A tool that assists with a draft does not perform the entire job or assume responsibility for the service.

List the tasks before discussing automation. Identify which require physical presence, professional credentials, relationships, confidential information, specialist judgment, or organizational authority. Note where a draft or information-retrieval aid might be relevant and where a different process is needed.

Keep the language precise. “The tool prepared a draft outline” is an observation. “The role can be replaced” is a much broader conclusion that needs additional evidence and organizational decisions. Do not turn one into the other.

Create a task inventory with employees

Ask staff how the work actually happens, including exceptions and informal coordination. Process documents may omit the steps people take to resolve missing information, explain an unusual case, or check whether something is appropriate. Those steps matter when evaluating a proposed workflow.

Record the input, action, output, frequency, dependencies, and person responsible. Identify the system of record and who can approve changes. Include work that occurs only during a reporting period, event, audit, or other less frequent cycle.

Discuss the purpose of the review clearly. If the organization is evaluating a particular tool, explain the task and decision involved. Avoid asking employees to describe their work under a vague promise that nothing will change or a speculative claim that their roles will disappear.

Program team discussing the tasks and decisions involved in their work
Understand the full workflow before drawing conclusions about a job.

Identify the type of assistance being considered

Different uses imply different responsibilities
Use Example Question to resolve
Drafting Prepare an outline from approved notes. Who verifies meaning and facts?
Retrieval Find an approved policy. Is the source current and authorized?
Analysis support Organize a dataset for review. Are the method, data, and interpretation appropriate?
Action Update a record in a connected system. Who authorizes and checks the change?
Consequential decision A decision affecting employment or service access. What specialist review and accountable process applies?

Use the nonprofit AI workflow guide to compare tools and pilots. The current AI recruiting guide addresses connected knowledge, research, and administrative assistance in hiring.

Test a workflow before assuming its impact

Select a bounded task with approved or fictional information. Define the result you need, the reviewer, and the questions the test should answer. Include incomplete input, conflicting details, and a case that should be escalated instead of completed automatically.

Compare the output with the source and record corrections. Include the time spent preparing input, checking the result, and resolving errors. First-draft speed is only one part of the work. A workflow that produces fluent text can still require substantial verification.

Keep observed results separate from predictions. A pilot may support a narrow use while leaving other tasks unchanged. Document the conditions of the test and the limits of what you learned. Reassess after material changes in the tool or process.

Consider the people affected by the workflow

Employees, applicants, volunteers, donors, participants, and partner organizations may experience a change differently. Ask whose information is used, who receives the output, and who has a way to question or correct it. Include the people closest to the service in the discussion where appropriate.

A public-facing assistant needs an accurate purpose and a way to reach a person when the question is outside scope. A system used internally still needs clear ownership. Do not assume that a conversational interface can safely handle urgent, clinical, legal, or sensitive service requests.

Review accessibility in both the tool and the resulting process. Consider whether someone can use another channel, understand the output, and raise a barrier. A workflow should not depend on one communication style or on access to a particular personal device.

Plan learning around capabilities

Useful learning can include source verification, task definition, data handling, editing, documentation, and recognizing limitations. These capabilities relate to professional judgment as well as tool operation. Someone who can identify an unsupported statement may contribute more to a workflow than someone who can produce a polished draft quickly.

Choose learning that matches the person’s actual responsibilities. Communications staff may need practice checking generated copy, while a system administrator may need to understand access and integration controls. Do not assign every employee an undefined requirement to become an AI specialist.

Give staff time, examples, and a place to ask questions. Clarify permitted uses and information before expecting experimentation. Review whether learning activities are accessible and feasible alongside the work people already perform.

Reviewing a role? Describe the tasks and capabilities clearly, then use the hiring resources and posting plans. Candidates can compare current openings.

Update responsibilities explicitly

If a pilot changes the work, review the role description, supervision, training, and resources. Identify who verifies output, maintains source material, handles exceptions, and approves changes. Those responsibilities take time and should not remain invisible.

Distinguish operating a tool from governing it. A person using a drafting assistant may not have the authority to approve a new connected app. A person maintaining a knowledge source may not decide how information can be shared. Explain the boundaries and escalation route.

Keep employment decisions within the organization’s appropriate process. Review applicable requirements and communicate changes through accountable people. Avoid making promises about job security or advancement that the organization has not approved and cannot substantiate.

Professional examining reports and digital information about a proposed workflow
Record evidence from a test alongside its limitations.

Read AI and employment research carefully

Check what a study measures: exposure of tasks to a technology, reported adoption, observed productivity, employer intentions, or actual employment. These are different findings. A model of technical potential does not establish that organizations will adopt the technology or change staffing in the same way.

Read the date, geography, occupations, sample, and assumptions. Avoid turning a global estimate into a local nonprofit prediction. An average can also hide differences among functions, organizations, and workers. State those limits when using research in an internal plan.

NIST’s AI Risk Management Framework offers voluntary risk-management context. BLS employment projections provide occupation-based labor-market information. Neither source determines the outcome of a particular technology project or an individual career decision.

Review privacy and decision responsibilities

Identify the data needed and who is authorized to provide it. Use the organization’s approved review for personal, confidential, or sensitive information. Check access, retention, deletion, exports, and connected systems before using real operational records in a tool.

Employment tools require separate attention from general writing assistance. The EEOC’s AI and disability resources provide a starting point for reviewing employment-related concerns. Obtain appropriate advice for the proposed use rather than assuming a vendor’s description answers every requirement.

Keep a responsible person and a route for questions visible. An automated step should not leave employees or service users unable to understand who owns a decision. Document how an error is identified, corrected, and communicated.

For job seekers: Describe experience concretely

Explain the work you performed, the tool or system used, the information involved, and the review you completed. A specific example of finding an error or organizing a task is more informative than a broad claim to be future-ready.

Distinguish introductory learning, supervised practice, and independent responsibility. Keep credentials and proficiency claims accurate. If a role includes an unfamiliar system, ask what level of capability is required and what training is available.

Choose current vacancies based on responsibilities and employment conditions rather than a prediction about which job will survive. Use the resume guide to communicate relevant experience and the job-search guide to evaluate opportunities.

Example: A reporting workflow changes

Imagine a fictional program team that prepares monthly narrative summaries from approved aggregate data. A pilot tool organizes notes into an outline. Staff still verify the underlying information, explain unusual cases, select appropriate conclusions, and approve the final report.

The team records whether the outline is useful and how much correction it needs. If the tool invents a program result, that is documented as an error rather than accepted because the wording is persuasive. The next decision concerns the defined outlining task, not the entire program role.

If the team adopts that narrow use, it assigns a source owner, reviewer, and escalation route. It also reviews workload so checking the output is recognized. The example illustrates a decision process without predicting savings or staffing changes.

Foundation colleagues discussing work responsibilities during a meeting
Keep review, ownership, and communication part of a technology decision.

Use a workforce review meeting to make the next decision

Bring the task inventory, pilot observations, proposed responsibilities, and unanswered questions. Start by describing the work people need to accomplish. Ask staff to identify exceptions or dependencies the pilot did not cover. Keep a record of the discussion that distinguishes facts, assumptions, and decisions.

Next, compare the proposed workflow with the current one. What information must be prepared? Which steps change? Who checks the result? Where can an error affect another team or a member of the public? Identify the practical support needed before expanding access.

Review the implications for workload and learning. A tool may reduce one task while adding preparation, verification, or maintenance elsewhere. Assigning those tasks explicitly provides a clearer picture than counting only the activity that changed. Discuss how staff can raise concerns and how the organization will respond.

Finish with a narrow decision and an owner. You might approve another fictional-data test, permit a defined drafting use, seek specialist advice, or stop the proposal. State the conditions and the next review point. An accountable decision is more useful than a general instruction to adopt AI quickly.

Communicate with the people doing the work

Explain what has been decided and what remains open. Give staff the approved instructions and a contact for questions. If the pilot changes, communicate the change rather than assuming everyone will notice a revised setting or a new system permission.

Make room for practical feedback. Someone using the workflow daily may identify confusing output, repeated corrections, or an exception the project team missed. Treat that information as evidence for the next review. Avoid framing a report of problems as resistance to technology.

Recognize that confidence and competence develop differently. Some staff may be comfortable experimenting while others want a clearer example or supervised practice. Provide learning related to the actual task and assess the work rather than enthusiasm for a product. Keep the conversation connected to responsibilities, evidence, and support.

Keep the review record and approved instructions available so new colleagues can understand the workflow, its limits, and the responsibilities attached to it.

Questions about AI and nonprofit jobs

Does a task being technically possible mean a job will disappear?

No. Jobs combine tasks, responsibilities, relationships, and organizational decisions. Examine the actual workflow and evidence.

Should every role require AI experience?

Describe the capabilities needed for the work. Different roles may need different tools, judgment, or training.

Can a pilot establish organization-wide results?

A pilot provides evidence about the tested use and conditions. Broader adoption needs a separate assessment.

How can staff prepare for changes?

Learn relevant task definition, source checking, data handling, and review practices, and ask how responsibilities and support are being defined.

Connect workforce decisions to the work

Employers can define roles clearly and review posting options. Candidates can explore actual openings and prepare examples of relevant experience.

View employer optionsSearch nonprofit jobs

FOUNDATION LIST CAREERS

Continue Your Mission-Driven Career

Explore more openings from mission-driven employers.

Explore All Nonprofit Jobs