Chapter 8: Generative Artificial Intelligence

Table of Contents

1. Understanding Generative AI Tools
2. Getting Started with AI Prompts
3. AI for Adult Learners
4. AI for Career Navigation
5. AI for Administrators and Support Staff
6. Data Privacy, Ethics, and Use Policies
7. Use Policies for Programs and Classrooms
8. Frameworks for Responsible AI Integration
9. AI Tool Overview
10. Emerging AI Features
11. AI Tools for Adult Education

Artificial intelligence (AI) is increasingly becoming part of adult education, influencing how programs approach teaching, administration, and learner services. Some AI applications are already part of daily life, such as predictive text on phones, navigation apps, music recommendations, and fraud detection in banking. However, this chapter focuses on generative AI, which refers to tools that create new content such as text, images, audio, or video in response to prompts.[1] Focusing on ways generative AI can change how educators plan, deliver, and support learning, this chapter considers both opportunities and risks. The emphasis is on approaches that can adapt as tools continue to evolve, rather than on specific products that may change quickly.

Understanding Generative AI Tools

Before exploring specific applications, it is useful to understand the basics of how generative AI works. This context can help educators evaluate both its potential uses and its limitations.

Generative AI systems create new content in response to prompts. Text-based systems, often called large language models, are trained on extensive collections of data that include books, websites, articles, and other sources. Rather than “knowing” facts, they generate responses by predicting likely words or sequences based on patterns in their training data.

This process is similar to predictive text features in email or messaging applications, which suggest the next word or phrase based on prior use. Generative AI carries out a comparable function at a larger scale, drawing on extensive datasets and more advanced computing models to produce extended and contextually relevant responses.

Newer tools extend beyond text. Many can accept voice input, analyze images or videos, read PDFs, and generate visuals from text prompts. These systems work by combining language models with image and audio models, allowing the AI to interpret and connect information across formats. For example, a learner might upload a chart and ask the tool to explain its trends, or describe a scene in words and receive a generated image in return.

Because these tools are trained to predict patterns in language, they can produce answers that appear fluent and convincing even when the information is inaccurate. This can lead people to assume a response is correct when it is not. AI may fabricate details (sometimes called a “hallucination”) or reflect biases from its training data. Human judgment remains essential, including verifying information and recognizing that responses are shaped by existing content. That content may underrepresent diverse perspectives, overlook minority viewpoints, or reinforce stereotypes.

That’s why it’s essential for educators to review AI-generated content. Research in higher education supports this point, showing that AI is most effective as a drafting tool when paired with careful human editing and contextualization, rather than as an end-to-end solution.[2]

The next section describes how AI can be used across different roles and provides examples that illustrate its application in adult education settings.

Getting Started with AI Prompts

An essential skill for using generative AI is learning how to write clear prompts. General or incomplete prompts often lead to vague results, while well-structured prompts can generate usable output that requires minimal editing.

Prompt formulas are structured templates that guide the AI with specific elements such as:

• Persona: the role or perspective the AI should take (for example, “You are an ESL teacher…”).
• Context: background information about the learners or setting.
• Task: the specific request (lesson plan, summary, explanation, etc.).
• Exemplar: an example to follow or a model to use.
• Format: the structure of the output (outline, table, email, etc.).
• Tone: the style or voice (formal, friendly, plain language, etc.).

For example:

• Before (general): “Can you help make an intermediate level ESL lesson plan?”
• After (structured): “You are an ESL teacher at a community-based adult school (persona). The learners are mostly Spanish-speaking adults at CASAS level 2 (context). Create a 60-minute ESL lesson plan that focuses on filling out a job application (task). Use the CASAS employment forms unit as a model (exemplar). Present it as a simple outline with objectives, activities, and materials (format). Write in a clear and professional voice (tone).”

The second prompt includes the level of detail needed for the AI to generate a lesson plan that is specific, potentially usable with editing, and aligned with adult learner needs.

Additionally, instead of trying to write a perfect prompt on the first attempt, users can start by asking the AI to ask them questions. This can help clarify goals, gather missing details, and create a stronger foundation before drafting any output. For example, an administrator preparing a grant proposal could begin by prompting, “Ask me questions to help define our program goals and data needs for a technology integration plan.” The AI’s follow-up questions can surface missing information and help structure the planning process more efficiently.

When available, attaching or uploading relevant materials (such as a syllabus, lesson plan, or dataset) can also help AI tailor its responses to the adult education setting. For example: “Using the uploaded syllabus, create three short reflection questions for learners.”

Providing this context helps the AI generate outputs that are aligned with local curriculum, learner goals, or reporting formats. Using these strategies, educators can move from generic prompts to specific, context-aware conversations that produce more accurate and useful results.

AI for Instruction

Some AI tools are advertised to support lesson design, classroom activities, and assessment. While they do not replace educator expertise, they might be used to complement instructional planning and provide new ways to create, organize, or adjust learning materials.

Lesson and activity generation: AI can draft lesson plans, warm-ups, or practice questions prompted to align with objectives. For instance, an instructor preparing a unit on financial literacy could request practice scenarios involving budgeting or paycheck deductions.

• Example prompt: “Create a list of five role-play scenarios for an adult education math class on personal finance. Each scenario should involve calculating income, expenses, or savings in a real-world context.”

Considerations for Reviewing AI-Generated Lessons: Some lesson generators reflect built-in pedagogical biases, such as teacher-centered delivery, rigid sequencing, or assumptions drawn from K–12 contexts. These patterns may not align with adult learning theories that emphasize relevance, flexibility, and learner agency. Reviewing and adapting AI-generated lessons helps ensure that pacing, tone, and structure reflect adult learners’ goals, local contexts, and program outcomes.[3]

Content adaptation and differentiation: Instructors can ask AI to simplify, summarize, or reformat texts in an effort to make materials more accessible. For example, a workplace safety article could be rewritten in plainer language, paired with a glossary of key terms, or followed by comprehension questions. This allows learners at different proficiency levels to engage with the same core content.

• Example prompt: “You are an assistant helping prepare materials for an ESL class. Rewrite this workplace safety article at a 6th-grade reading level, add a glossary of five key terms, and generate three multiple-choice comprehension questions with answers.”

Writing support and feedback: AI can suggest edits to learner writing or model how revision works. For example, with the right prompt, pasting a paragraph into an AI tool can generate grammar corrections and explanations. Educators can adapt these suggestions into a mini-lesson, a peer review activity, or targeted feedback.

• Example prompt: “Review the following student paragraph. Highlight any grammar or punctuation issues, explain how to fix them in plain language, and suggest one way to make the writing clearer. Text: [paste paragraph].”

Classroom Activities: AI can generate both strong and weak versions of an assignment to support critical evaluation. For example, in a job readiness unit, an instructor could provide a job description and ask AI to produce two sets of interview responses, one professional and specific and another vague or incomplete. Learners then analyze and compare the responses.[4]

• Example prompt: “Using this job description: [paste text], write two versions of answers to five common interview questions. The first set should be strong, specific, and professional. The second set should be vague or incomplete. Present both sets side by side.”

Formative assessment support: AI can generate practice quizzes, exit tickets, or reflective prompts to help check understanding. It can also produce sample answers or explanations to support analysis of learner work.

• Example prompt: “Generate five short-answer questions that check comprehension of this reading passage. Provide sample correct and incorrect answers written at a 7th-grade reading level. Text: [paste passage].”

Because evidence-based best practices for AI in instruction are still emerging, its use remains largely experimental. When applied thoughtfully, AI can help draft materials, sample activities, or practice assessments.

AI for Adult Learners

Developing AI literacy is now part of digital literacy for adult learners. Learners benefit not only from using AI tools but also from understanding how they work and how to use them critically and responsibly. Applications include clarifying complex concepts, practicing communication skills, preparing for assessments, and researching or planning career pathways. These activities can take place during class, as additional practice outside lessons, or while applying skills in workplace and community contexts.

Classroom discussions or reflection activities can help learners evaluate AI-generated information, recognize its limits, and consider appropriate uses. The AILit Framework provides a resource for defining and practicing these skills, organized around four domains: Engage, Create, Manage, and Design. Together, these concepts build a foundation for confident and informed use of AI in learning and daily life.

The examples below illustrate how these skills can be applied in classroom and independent learning contexts.

On-demand explanations and tutoring: AI can write about concepts in plain language, provide examples, and answer follow-up questions. For example, a learner preparing for the GED math test might prompt a step-by-step explanation of the triangle area formula from the GED formula sheet, then request practice problems. Learners can also prompt AI to adjust explanations for different reading levels, connect examples to career goals, or show multiple approaches to the same problem.

• Example prompt: “You are a math tutor helping a student prepare for the GED. Explain the origin of the triangle area formula from the GED formula sheet in plain language. Then, create three practice problems with answers, using examples related to construction or carpentry.”

Practicing conversation and language skills: Learners can use AI to simulate everyday conversations, receive feedback on grammar and vocabulary, and explore clearer phrasing options. English language learners practicing for a workplace might rehearse interactions such as greeting customers, explaining a product, or responding to a supervisor. Some tools include voice input and output, which could be used to practice pronunciation and listening comprehension.

• Example prompt: “You are a conversation partner for an English learner practicing customer service. Role-play a short dialogue between a customer and a cashier at a grocery store. After the conversation, list any grammar or vocabulary errors and suggest clearer ways to communicate.”

Study aids and skill practice: Learners preparing for academic tests, high school equivalency exams, citizenship interviews, or professional certifications can use AI to generate practice materials. These might include questions with answers, flashcards, plain-language summaries of complex topics, or definitions of unfamiliar vocabulary. Learners must still review and verify the accuracy of the information it provides.

• Example prompt: “You are a study assistant. Create 10 practice questions for the U.S. citizenship test focused on U.S. history, with correct answers. Then, write a short summary of the main points covered, using plain language at a 7th-grade reading level.”

Career and life skills development: AI tools can help users prepare resumés, edit cover letters, and practice interview questions with feedback. They can also assist in drafting professional emails, explaining workplace or legal documents in plain language, and outlining steps for common life tasks. For learners, practicing these activities with AI can provide opportunities to build confidence in a low-stakes environment before applying them in authentic contexts.

• Example prompt: “You are a career coach helping someone tailor their resumé for a specific job. First, review this job description: [paste job description]. Next, review this resumé: [paste resumé]. Suggest specific changes to align the resumé with the posting, focusing on relevant ATS keywords, skills, and measurable accomplishments. Recommend any sections to reorganize for clarity, and flag information that could be shortened or removed. End by drafting a short summary statement for the top of the resumé that highlights the applicant’s fit for the role.”

Role of the Educator

Educators can model how to use AI effectively and responsibly by demonstrating clear prompting, reviewing AI-generated responses with learners, and discussing accuracy and usefulness. For example, a class might pose the same question to AI in several ways, compare the responses, and evaluate which is most accurate or clear.

To facilitate this process, instructors can display prompts and results on a shared screen, invite learners to suggest revisions, and discuss how changes affect the outcome. Small groups can then repeat the exercise with new topics and share what they learned.

Practical classroom activities can reinforce these habits and help learners engage critically with AI-generated content. Examples include:

• Drafting a professional email independently, then generating another version with AI and comparing the two for clarity and tone.
• Working in pairs to refine a prompt until the AI produces a clear explanation of a concept.
• Summarize a short article or video independently, then generate an AI summary of the same material. Compare which details each version includes or omits, and discuss how those choices affect meaning and credibility.

These activities work best when educators guide reflection after each task by asking what was helpful, what needed correction, and how learners might apply similar strategies in their own study or work contexts, emphasizing that AI output is a starting point for review and discussion, not a finished product.[5]

Clear expectations can help learners understand when and how AI can be used appropriately. Like a calculator that supports problem-solving without replacing the need for conceptual understanding—or a translation tool that assists with language comprehension but cannot capture full meaning or tone—AI can help with drafting, brainstorming, or language practice while the learner remains responsible for reasoning and accuracy. Educators can clarify when AI use is encouraged, when original work is required, and how to acknowledge AI assistance. These boundaries help maintain academic integrity while supporting responsible exploration of AI as a learning aid.

By positioning AI as one of several tools and openly discussing its strengths and limitations, educators can help learners build the skills to use AI in ways that are accurate, ethical, and aligned with their goals.

AI for Career Navigation

AI tools can enhance career navigation activities in adult education, supporting both staff and learners. From exploring career options to preparing application materials, these tools can help connect classroom learning with practical workforce opportunities.

Labor market information: AI can generate summaries of local job opportunities by drawing on current labor market data and training program information. It can identify in-demand jobs, list common entry titles, note typical wages, and describe the training or certifications that employers require. Outputs can be formatted as comparison tables, bilingual handouts, or short summaries written at a specified reading level.

• Example prompt: “You are an assistant helping an adult education program prepare career information for learners. Using publicly available labor market data for [region] from sources such as the Bureau of Labor Statistics, ONET, and the state labor market dashboard, create a table of five entry-level jobs projected to grow in the next five years. For each job, include: (1) typical job titles, (2) median hourly wage in [region], (3) common training or certification requirements, and (4) one public training provider in [region] that offers the required training. Then, write a 150-word summary explaining these jobs in plain language at a 7th-grade reading level. Make sure all information is current and include links to sources.”

Resumé and cover letter support: AI can assist in drafting or refining resumés and cover letters by aligning documents to a specific job posting. This can provide learners with a strong starting point, though best practices still involve human coaching to ensure final documents are accurate and authentic. AI can identify relevant keywords, reorganize sections for clarity, and adjust language to better match employer expectations. Because personal data should not be entered into public AI tools, placeholders or anonymized profiles (for example, “[Student Name]”) should be used.

• Example prompt: “You are an assistant helping an adult education program support a student applying for a job. Using the anonymized resumé text and the job posting below, suggest specific improvements to align the resumé with the requirements of the posting. Highlight transferable skills, recommend keywords to include, and adjust section order if needed. Then, draft a three-sentence cover letter that connects the student’s experience to the role. Use placeholders for all personal details. Resumé text: [paste text]. Job posting: [paste text].”

Instructor Tip: Triple quotation marks (""") can separate an AI prompt’s directions from the material being analyzed or rewritten. For example, place the job posting inside the triple quotes and write the prompt outside them. This makes it clear that the AI should process only the job posting, not the instructions themselves.

Interview preparation: AI can be used to simulate job interviews by posing questions and offering feedback on responses. Some tools allow interactive role-play where questions are asked one at a time, followed by comments or suggestions for improvement. This can give learners opportunities to practice with common interview formats and receive targeted feedback in a low-stakes setting. While AI cannot replace the nuance and empathy of a human coach, it can supplement existing interview preparation activities.

• Example prompt: “You are an interview coach helping prepare a learner for a retail cashier position. Ask five common interview questions, one at a time. After each answer, provide one constructive suggestion to improve the response and one tip for effective communication. Focus on clarity, relevant examples, and professional tone.”

Tool feature: Career Dreamer

Career Dreamer is a free tool from Grow with Google that uses AI to help learners explore career options based on their skills, experiences, and interests. It prompts learners to describe their background and then generates a short “Career Identity Statement” that summarizes key strengths in plain language. Drawing on U.S. labor market data from sources such as the Bureau of Labor Statistics and Lightcast, the tool suggests possible career paths and identifies typical training needs. It also connects to Google’s Gemini AI for resumé and cover letter drafting. Career Dreamer runs in a web browser, stores information locally on the user’s device, and is currently usable without an account (as of 2025).

AI for Administrators and Support Staff

AI can streamline many administrative and support functions in adult education. It can draft communications, organize data, summarize reports, and generate outreach materials, which could help programs save time for tasks that depend on human judgment, relationships, and decision-making.[6,7]

Research from statewide professional development initiatives also shows that AI can assist with reviewing written reflections and lesson artifacts that educators submit to earn micro-credentials. In one pilot,[8] an AI-supported scoring system analyzed these submissions using program rubrics, reducing review time by about one-third while maintaining consistent results. This could make it easier to scale professional learning opportunities and report outcomes without increasing staff workload or reducing quality.

The examples below illustrate other general ideas for AI to support administrative and support roles.

Communication and Outreach: Programs often need clear, timely messages for learners, partners, and stakeholders. Drafting these from scratch takes time, but AI can generate initial text for staff to refine.

• Example prompt: “Draft a short, friendly email reminding learners about the upcoming registration deadline. Include the documents they should bring. Provide versions in English and Spanish.”
• For a digital navigator: “Create step-by-step instructions for a Spanish-speaking learner to install Zoom on a smartphone.”

AI can also suggest alternate formats, such as adapting an announcement into a social media post or creating content for a printed flyer.

Reporting and Data Summaries: Quarterly and annual reports often require turning enrollment, outcome, and program data into a clear summary for funders, consortium partners, or the public. AI can help by reviewing anonymized data and drafting a readable summary that staff can then refine.

• Example prompt: “Review the enrollment and outcome data provided below and draft a short summary of this quarter’s results for inclusion in a WIOA report. Highlight key trends such as increases in enrollment or measurable skill gains, and write in a clear, professional tone suitable for program reporting. Data: [paste anonymized table or data].”

Translation and Accessibility: Many programs serve multilingual communities. While professional translation should be used for formal documents, AI can provide informal translations for everyday communication or quick drafts of materials that need to be shared in multiple languages.

• Example prompt: “Translate the following attendance policy into Arabic and write it in a clear, conversational tone appropriate for adult learners. Keep key terms consistent and avoid literal word-for-word translation. Text: [paste policy].”

Administrator Tip: Use Plain Language Standards
The Plain Writing Act requires agencies to communicate in clear, accessible language. The same principles can guide adult education programs in adapting AI-generated content for learners. Resources and examples are available at plainlanguage.gov.

• Example prompt: “Revise the following attendance policy to meet Plain Language Act standards. Write it so that it is easy for adult learners and staff to understand. Text: [paste policy].”

Scheduling and Routine Questions: AI chatbots and scheduling tools can handle common questions and basic appointment or registration scheduling. When connected to a program’s trusted information sources, like program documentation, a chatbot can respond to questions such as “When do ESL classes start?” or “How do I register?” without staff needing to reply manually.

• Example prompt: “Using the program information below, create a list of 10 clear questions and answers that adult learners commonly ask about ESL classes. [paste text].”

Grant Writing and Strategic Reports: AI can help staff organize and draft complex documents such as grant narratives, annual reports, and strategic plans. For example, an administrator preparing a WIOA or CAEP grant can upload planning notes, enrollment data, and key objectives, then ask the AI to generate a structured narrative that meets funder requirements. The AI can outline sections, suggest language for measurable outcomes, and align program goals with federal or state priorities. Staff can then focus on reviewing, editing, and confirming accuracy before submission.

• Example prompt: “Using the notes and requirements from the grant application below, draft a two-page narrative describing how our adult education program will expand digital literacy instruction. Include specific goals, expected outcomes, and connections to WIOA and CAEP priorities. [paste notes and requirements].”

Administrator Tip: For complex projects such as multi-year grants or strategic plans, AI can be prompted to ask clarifying questions before drafting. For example: “What questions do you have for me that will help you create a draft project plan?” This creates an iterative process in which the system generates questions, staff provide answers, and, over several rounds, a more complete and tailored plan emerges.

Data Analysis and Trends: AI can help identify patterns in enrollment, attendance, persistence, and assessment results. For example, it might highlight declining enrollment in a program area, identify links between attendance and test gains, or compare outcomes across cohorts.

• Example prompt: “Using the attached datasets (attendance, CASAS post-test results, and course completion rates for the past three program years), identify the instructional areas with the highest and lowest academic progress as measured by skill gains. Highlight any correlations between instructor, attendance patterns, and test score improvements. Based on these findings, suggest three specific areas where curriculum adjustments or targeted interventions might improve learner outcomes. Include a short summary that program leadership could use to set priorities for professional development, resource allocation, and partnership development.”

Resource and Partnership Mapping: AI can compile and format information about local resources such as childcare, transportation, or training programs. This helps maintain updated referral lists for staff and learners.

• Example prompt: “Using the following list of verified local organizations, create a table showing each organization’s name, service type, eligibility requirements, and contact information: [paste list or links].”

Administrator tip: Programs conducting resource mapping or labor market analysis can benefit from “deep research” modes in some AI tools (ChatGPT, Gemini). These modes search across multiple sources—such as state workforce dashboards, local college catalogs, or nonprofit service directories—and organize the findings into a usable format. For example, an administrator might ask the AI to identify childcare providers within a county that serve adult learners or to compare entry-level health care training options across nearby colleges. The tool compiles links, contact details, and short descriptions, which staff can then verify and add to referral lists or partnership directories.

Data Privacy, Ethics, and Use Policies

Responsible implementation of AI for adult education programs involves protecting data privacy, ensuring fairness and accuracy, and setting clear expectations for appropriate use. This section outlines key practices to support safe and ethical adoption of AI across classrooms and operations.

Data Privacy and Consent

Earlier sections introduced privacy as a core consideration when using AI tools. In practice, these concerns center on what information is shared, where it is stored, and who has access. Most free or public AI services store user inputs and may reuse them for future model training unless otherwise specified.

For educators and administrators, two key considerations apply:

• Avoid inputting sensitive learner information. Personally identifiable information (PII)—such as names, test results, or contact details—should never be placed in public AI systems. When examples are needed, use placeholders or fictional learner profiles instead.
• Review the tool’s data policy and settings. Education-specific or enterprise versions often provide stronger protections, including data-use restrictions or storage opt-outs. For example, Microsoft 365 Copilot specifies that prompts and responses are not used to train foundation models, while ChatGPT free accounts indicate that user content may be used for model training.

Programs should balance the convenience of free tools with the need for privacy and compliance, choosing platforms that best protect learner and program data.

When learners use AI tools directly, such as asking for resumé feedback or practicing English conversation, they should understand how their inputs may be used. Programs can support this by vetting tools and preparing simple guidance documents. Key points might include:

• Advice to not share personal contact information or ID numbers in the AI.
• Warning that conversations with the AI may be stored on company servers.
• Steps for deleting or turning off history, managing data permissions, or reviewing privacy options when available.

Copyright Considerations

AI tools raise copyright issues as well as privacy concerns. U.S. copyright law does not recognize AI as an author, so material created entirely by AI is not protected by copyright.[9] For educators and learners, the more relevant issue is how to use copyrighted materials responsibly and how to acknowledge AI contributions transparently.

• Attribution and transparency. When AI is used to generate or adapt text, images, or media, noting that use promotes clarity and academic honesty. For example, a learner might include a short statement such as, “Portions of this report were drafted with AI assistance and then edited by me.” Instructors can model this practice by sharing when materials were created with AI support.
• Using copyrighted materials with AI. Full copyrighted works, such as textbooks, journal articles, or courseware, should not be pasted into public AI tools, since doing so may violate license terms or exceed fair use. Using short excerpts or summaries for instructional purposes is generally allowed, particularly in nonprofit education. When in doubt, review whether the material is openly licensed (such as Creative Commons) or obtain permission for use.
• Human authorship. AI-generated material can be freely shared and adapted, but only content that includes substantial and creative human contribution qualifies for copyright protection. This distinction matters most when programs create official materials, publications, or course content. Human editing and authorship should be evident in all shared work to meet these requirements.
• Platform licensing. Programs should review the licensing terms of AI platforms, as some providers reserve rights to reuse user inputs and outputs. Checking these terms helps avoid unintentional sharing of instructional materials or sensitive data.

Instructor Tip: Many AI platforms allow users to share their chat threads via a link. This feature can support classroom activities that build critical thinking and transparency around AI use. For example, instructors can:

• Ask learners to submit their AI conversation link along with an assignment to show how they developed their ideas or refined drafts.
• Have students exchange and review one another’s AI threads to compare prompts, evaluate reasoning, and discuss differences in how each person guided the tool.
• Use shared threads in group discussions to highlight effective questioning, bias detection, or fact-checking strategies.

These practices help learners treat AI as a collaborative workspace for reflection and analysis rather than a shortcut for completing tasks.

Bias and Misinformation

AI tools are trained on large collections of text and images from the internet and other sources. This data can include inaccuracies, outdated information, and cultural, racial, or gender biases. As a result, AI-generated content can reflect or amplify misinformation and stereotypes.[10,11]

For example, an AI might:

• Draft a job description that favors one demographic group over another
• Summarize labor market information that is no longer current
• Present opinions on a policy issue as fact without acknowledging multiple perspectives
• Generate images that reinforce stereotypes, such as showing only women as flight attendants or only men as CEOs

Because of these risks, AI outputs should be reviewed for both factual accuracy and bias. This includes checking claims against reliable sources, scanning for insensitive or exclusionary language, and considering perspectives that may be missing. These practices align with information literacy skills already targeted in adult education, such as evaluating websites and media for accuracy and balance. For instance, just as learners are encouraged to question whether information on YouTube or Wikipedia is current and reliable, the same critical lens should be applied to AI-generated content.

Some AI tools now provide sources automatically (for example, Perplexity) or can be prompted to cite them, making verification easier. Regardless of the tool, AI should be treated as another information source that requires evaluation.

• Example prompt: “Summarize the following text and provide a list of reliable sources to support each key claim. Include active links to government, nonprofit, or academic sites where available.”

Environmental Considerations

AI has environmental impacts in addition to instructional and ethical considerations. Training and operating AI systems requires electricity and water. Energy use contributes to carbon emissions and can strain power systems, and water is often needed to cool data centers, placing pressure on local supplies in drought-prone areas. Producing and disposing of hardware such as chips and servers also has environmental impacts.

As investments in renewable and carbon-neutral energy sources such as solar, wind, and nuclear expand[12], the environmental impact of AI may decrease, though this remains an evolving conversation and consideration.

For individual adult education programs, the impact of a few queries may seem small. However, at scale, with millions of users worldwide, the collective effect is more noticeable and has led to broader discussions about AI sustainability.

One way to respond is to treat AI as a purposeful resource rather than a default option. The When do I AI? framework (from World Education and TCALL) encourages educators to first identify what is needed (like lesson ideas, resources, media), then check what already exists (like colleagues, standards, curricula, open repositories). AI can then be applied more strategically, by adapting or extending existing resources instead of using it as the first step to generate new content.

Framing environmental impact as part of responsible technology use helps programs make informed decisions that balance efficiency, accuracy, and sustainability.

Use Policies for Programs and Classrooms

Local use policies can help programs set shared guidelines for how AI tools are introduced and managed. These policies support consistency, transparency, and trust across classrooms and operations. They do not need to be lengthy; even a short list of bullet points can clarify boundaries and responsibilities.

Programs developing or updating local policies can draw on existing models such as the ISTE Acceptable Use Policy Guide and the TeachAI Toolkit for Responsible AI Use. Both provide frameworks, templates, and guiding questions that can help tailor an approach suited to adult education settings.

More example policy items below can help programs decide what to include or adapt when developing local AI use policies.

Allowable Use

Authorized Tools: Use only AI tools approved by the program or consortium that meet data privacy and security standards (for example, Microsoft Copilot, Gemini for Education, or other tools managed under institutional licenses).

How to Share Data: Only share information that is already public or fully anonymized. Learner data protected by FERPA or other privacy laws should never be entered into public AI systems.

Permitted Application Areas: AI may be used to draft documents, plan lessons, summarize meeting notes, or analyze de-identified program data. All outputs should be reviewed for accuracy and relevance before use.

Prohibited Use

• Unapproved Tools: Do not use personal or free AI accounts (such as public ChatGPT) for materials containing learner or staff data.
• Sensitive or Confidential Information: Never upload names, test scores, case notes, or personally identifiable information into AI systems not covered by institutional agreements.
• Assessment and Grading: AI should not be used to assign grades, make enrollment decisions, or evaluate learner progress without human review.
• Inappropriate or Misleading Content: AI tools must not be used to create false information, impersonate others, or generate discriminatory or harmful material.

Additional Guidance

• Transparency: When AI is used to draft, summarize, or translate materials, note its contribution clearly.
• Verification: All AI-generated materials should be checked for accuracy, bias, and cultural relevance before sharing.
• Training and Support: Programs are encouraged to provide staff orientation on approved tools, data privacy, and responsible use.
• Questions and Updates: Policies should be reviewed regularly as AI capabilities and institutional guidance evolve.

In classrooms, some educators establish norms collaboratively with learners. For example, an ESL class might agree, “AI translators can be used to draft stories, but learners must review and revise the translation,” or “AI will not be used during in-class timed essays.” Engaging learners in setting norms can increase understanding of the rationale behind rules.

Policies should be treated as living documents that evolve alongside new technologies and program needs. Rather than relying on AI detection tools, which have shown inconsistent accuracy,[13] programs can focus on clear communication, periodic policy reviews, and ongoing staff training. A proactive, transparent approach helps maintain program integrity and build trust among staff and learners as AI tools continue to develop.

Frameworks for Responsible AI Integration

Some established literacy and integration frameworks also outline key dimensions of AI literacy and offer examples of instructional and ethical considerations.

Research shows that AI literacy for adults is still emerging, with limited consensus on definitions and competencies. A review of 30 studies found broad agreement that AI literacy focuses less on programming and more on understanding, using, and critically reflecting on AI in daily and professional life.[14,15]

Two frameworks especially relevant to educators include:

• World Education AI Integration Framework: This framework offers guidance for integrating AI into adult education and workforce programs across six dimensions. Each dimension includes guiding questions, resources, and scenarios to support both technical and human-centered decision-making. The framework also includes an AI Tool Evaluation Rubric that can be used to assess suitability, accessibility, and alignment before adoption.
• AILit Framework: This framework organizes AI literacy into four domains—Engage, Create, Manage, and Design—to help educators, leaders, and developers build shared language and practical approaches to AI use.

Programs can use these frameworks to:[16]

• Identify staff training needs in AI concepts, privacy, and ethics
• Inform learner-facing AI literacy activities
• Align adoption with recognized competencies and ethical principles
• Evaluate and refine AI-related practices over time

Evaluation remains a key gap. While some initiatives report positive learner outcomes, few use validated measures of AI literacy. Programs may benefit from adding short feedback surveys or reflection prompts to early pilots to track confidence, application, and impact over time.

The frameworks and principles outlined above can guide thoughtful, phased implementation. As with any new technology, successful adoption involves piloting tools, gathering feedback, and refining use over time. Programs can adapt these approaches to local goals and capacities, drawing on examples in earlier chapters for guidance on equity, access, and instructional design.

AI Tool Overview

When selecting AI tools, programs can start small and evaluate fit before wider adoption. Key questions include:

• Can the free version be tested before committing to a subscription?
• Does the privacy policy meet program data-sharing requirements, especially if learners will use the tool directly?
• Does it integrate with existing workflows, such as Google Workspace or Microsoft 365?
• Does it provide educator-developed templates or examples relevant to adult education?

The most effective tools are those that support program goals, align with existing systems, and can be sustained in daily practice.

The list below provides a broad overview of commonly used and emerging tools relevant to adult education. It can serve as a starting point for exploring options that align with program goals, technology environments, and learner needs.

One useful distinction separates general-purpose tools and ones designed for specialized educational tasks. General-purpose tools offer flexibility across many contexts, while specialized tools are streamlined for particular functions. Choosing between them involves weighing flexibility and customization against efficiency and focus.

• General-purpose chatbots. Tools such as ChatGPT, Claude, Gemini, CoPilot, Grok, and Perplexity can answer questions, draft text, summarize information, and provide explanations. They adapt to many contexts but require clear prompts for effective results.
• Specialized education tools. Examples include MagicSchool.ai and Diffit. These provide subject-specific generators, structured activities, or preset templates aligned with instructional goals. They can reduce preparation time and offer targeted support but are less flexible outside their intended uses.

In addition to standalone AI apps, many widely used platforms now include AI functions. YouTube is testing AI-generated topic summaries and quiz tools, while Google Workspace and Microsoft Office have added AI assistants for writing, analysis, and formatting. For adult education programs, this means AI may already be part of tools that staff and learners use daily.

Emerging AI Features

Several emerging AI capabilities may soon influence adult education and workforce contexts. These features are not yet commonly used, but awareness of these capabilities can help programs anticipate changes in the digital landscape and consider how they may align with priorities in administration, instruction, and learner support as the tools evolve.

• Image generators (Gemini Nano Banana, DALL·E): Tools that create images from text prompts, which may help staff produce draft visuals for outreach flyers, program websites, or instructional materials. Learners could use them to generate pictures that support vocabulary building or help illustrate ideas in writing or speaking activities.
• Vibe coding (Replit, Lovable): Tools that create apps or websites directly from natural language prompts. Programs might use these to prototype check-in apps or dashboards that visualize attendance trends.
• Knowledge-based assistants and Custom GPTs (ChatGPT Custom GPTs, Gemini Gems): Tools that can be trained with specific materials and answer questions based on that content. Programs can upload items such as handbooks, schedules, or FAQs, then embed the assistant on a website or learning platform so students can access clear, consistent, multilingual information at any time.
• Audio generators (ElevenLabs, Google Text-to-Speech): Tools that convert text into speech or create audio explanations. These may support listening practice or provide alternative formats for program information.
• Transcription and subtitle generators (YouTube Auto-Captions, Whisper Web): Tools that create transcripts, captions, or multilingual subtitles for recorded videos. This can improve accessibility for learners who benefit from text-supported media.
• Podcast or audio-summary generators (NotebookLM, Adobe Shasta): Tools that turn documents or notes into spoken summaries. These may provide flexible ways for learners to explore program documents or learning materials.
• Video generators (Veo3, Sora): Tools that create short videos from text prompts or storyboard descriptions. Programs might explore these for developing orientation videos or visual explanations of processes.
• AI agents (ChatGPT Agent): Tools that can complete multi-step tasks with limited supervision, such as scheduling meetings, sending reminders, or preparing recurring reports.

AI Tools for Adult Education

The following overview highlights commonly used and emerging AI tools relevant to adult education as of 2025. It is not exhaustive, as new tools and features continue to be released. The list is intended as a reference point for educators and staff to explore options that align with program goals, technology environments, and learner needs.

General-purpose AI Assistants

• ChatGPT (OpenAI): Provides strong general drafting and explanation capabilities that support lesson ideas, examples, and administrative writing.
• Claude (Anthropic): Offers the same core functions as ChatGPT, with the added ability to handle longer documents. Many educators also use Claude for structured analysis tasks or code interpretation because of its clear reasoning style.
• Gemini (Google AI): Shares similar drafting, summarization, and translation capabilities. Its distinguishing feature is integration with Google Workspace, which can streamline tasks for programs already working in Google Docs, Sheets, and Classroom.
• Perplexity: Provides many of the same generative features and includes real-time search with cited sources. This can help staff check facts, review policy updates, or gather current information from verified sites.
• Copilot (Microsoft): Offers comparable drafting and analysis abilities within Microsoft Word, Excel, Outlook, and Teams. Because it runs inside institution-managed accounts, it can support privacy and data-security requirements.
• Grok (xAI): Shares the same core drafting and explanation capabilities as the tools above. Its distinguishing feature is access to real-time information from public posts on X, which can help staff verify sources, check the currency of information, or review how topics are being discussed in public forums.

Specialized Education Tools

• MagicSchool.ai: Creates lesson plans, rubrics, quizzes, and scaffolds lesson topics.
• Diffit: Produces leveled reading materials and comprehension questions for multi-level ESL, HSE, or CTE classes.
• Khanmigo (Khan Academy): Provides AI tutoring for learners and instructional support for educators.
• Career Dreamer: Helps learners explore career options based on their skills, experiences, and interests, suggests possible career paths, identifies typical training needs, and drafts resumés and cover letters.

Language and Communication Tools

• Duolingo Max: Offers AI-driven role-play and feedback for language learning.
• ELSA Speak: Provides targeted pronunciation practice with real-time feedback.
• DeepL: Delivers translation for multilingual communication.
• Grammarly: Improves clarity and tone in learner writing or official documents.
• Paci AI: Designed for English language learners, offering personalized feedback on writing, grammar, and vocabulary with adaptive practice activities aligned to learner proficiency levels.

Visual, presentation, and video tools

• Canva: Creates flyers, slides, certificates, and visuals with templates and AI design support.
• SlidesAI: Turns text into slide presentations for instruction or training.
• Veed: Automates video editing, captioning, and subtitling for accessible learning content.

Research Tools

• NotebookLM (Google): Summarizes and answers questions from uploaded documents, policies, or lessons. It can also analyze YouTube videos to support discussion of key points and generate audio summaries or simple podcasts based on the materials provided.
• Scite: Locates research sources and shows how they have been cited to support evidence-based instruction.
• Elicit: Extracts relevant findings from research papers to inform curriculum design or grant writing.

1. https://www.gao.gov/assets/830/826491.pdf
2. https://digitalcommons.uri.edu/cba_facpubs/548/
3. http://osf.io/preprints/osf/zqjw5_v1
4. https://www.mdpi.com/2227-7102/13/10/998
5. https://www.mdpi.com/2227-7102/13/10/998
6. Artificial Intelligence and the Future of Teaching and Learning (PDF)
7. GENERATIVE AI AND THE FUTURE OF TEACHING AND LEARNING
8. https://www.learntechlib.org/primary/p/225199/
9. https://www.copyright.gov/ai/ai_policy_guidance.pdf
10. https://files.eric.ed.gov/fulltext/EJ1408776.pdf
11. https://www.ed.gov/sites/ed/files/documents/ai-report/ai-report.pdf
12. https://seia.org/blog/ai-leaders-are-pumping-billions-into-solar-storage/
13. https://arxiv.org/abs/2306.15666?utm
14. https://bera-journals.onlinelibrary.wiley.com/doi/abs/10.1002/rev3.3489
15. https://www.sciencedirect.com/science/article/pii/S2666920X2200056X?via%3Dihub
16. https://www.ed.gov/sites/ed/files/documents/ai-report/ai-report.pdf