Syllabus overview

The syllabus, readable here and available as a document on Canvas.

SIL 503, Technology for Social Good: Innovation and Ethics in Social Impact Leadership Fall 2026, 3 credits, In person in SSWB B760, Wednesdays 9:00 AM to 12:00 PM, September 9 to December 9, No prerequisites, Letter grades Instructor: Nari Yoo, PhD, nariyoo@umich.edu A course in the MA in Social Impact Leadership at the University of Michigan School of Social Work

Course description

This course examines the transformative role of technology in advancing social impact, with particular attention to current and emerging technological trends. Students will explore both the opportunities and ethical challenges posed by technology in social change work, including artificial intelligence, data systems, and digital tools for community engagement. The course emphasizes practical applications of technology in personal and organizational workflows, prompt engineering for AI tools, and understanding the community-level impacts of technological solutions. Students will develop principles for ethical use of technology that consider environmental sustainability, equitable access, and community well-being.

Using technology to deal with social problems is not new. Charity organization societies built case registries to sort families experiencing poverty. Hull House residents and W. E. B. Du Bois used maps and data portraits to document injustice. Today’s debates about AI belong to that longer history, and this course teaches both sides of it.

The course covers eight core areas: digital equity and the digital divide; technology ethics grounded in social values; data sovereignty and privacy; algorithmic bias and fairness; human-centered design; digital storytelling; technology accessibility; and ICT for development. Every technology receives the same interrogation, whether it harmed people or helped them, and the sociotechnical protocol is that interrogation. You use the tools yourself, every week, with disclosure and verification.

Learning objectives

By the end of the course you will be able to:

1. Analyze current and emerging technology trends and their applications for social impact work.
2. Apply AI and other technological tools effectively in personal, organizational, and administrative workflows, including prompt engineering techniques.
3. Assess the community-level impacts of technology interventions, including cost savings and accessibility considerations.
4. Develop and apply ethical principles for technology use that address environmental impact, equitable access, and community needs.
5. Design technology-informed solutions to social challenges that balance innovation with ethical considerations.

Grading

Full guidance, worked examples, and rubrics live on each assignment page.

Grading scale

Attendance

This course is built on in-class work, so attendance matters materially. You may miss up to two sessions for any reason without penalty. Three or more absences reduce the participation grade; more than four may lead to a conversation about withdrawing. Email me when you know you will be out. If you miss a lab, arrange to make it up; most labs can be completed asynchronously if we set it up ahead of time.

If you are ill, do not attend; no doctor’s note is needed. Email me and I will arrange access to the materials you missed.

Late work

Problem Inventory entries take no late credit; the deadlines are the Week 6 and Week 9-window checkpoints themselves. For the three major assignments you have five flex days to spend across the semester, each buying 24 penalty-free hours. Email me how many you are using; no explanation required. After flex days run out, work within 24 hours of the deadline earns up to 80% credit, and work more than 48 hours late is not accepted. Real emergencies are handled case by case; just reach out.

Bonus points

Office hours, up to 2 points on the final course grade. One point for coming early in the term to talk about your interests and goals, one for booking a meeting to discuss a major assignment. You do not need to bring a draft or send anything ahead; come with a question. Book at calendly.com/nari-yoo.

Event reflections, up to 3 points on the final course grade. I share relevant events on Canvas through the semester: talks, webinars, conferences, and community events. Attend one and post a short reflection or summary to the Canvas discussion within a week of the event, and it earns one point, up to three events.

Both kinds of bonus add to the final course grade rather than to any one component, so up to 5 points are available and they can offset a loss anywhere, including on a major assignment. Attending is optional and the reflection is marked on having been written rather than on how well.

AI policy

This course requires generative AI use, in the labs above all: you cannot evaluate a technology you have never handled. Two things come with that. Say where you used it; the class writes its own disclosure form together in Week 5. Check what it gives you before it enters your work, since an unverified AI error is your error. Hidden AI use is an academic integrity violation and disclosed use has no penalty. The judgment in your submissions should be yours, and the free tier of a general-purpose assistant covers every lab and assignment, so no graded work depends on a subscription.

Confident, well-formed, wrong output is the documented failure mode: one database has logged roughly 1,762 filings citing cases that do not exist, and KPMG withdrew a report on agentic AI after an audit found 5 of its 45 citations pointed to real sources. Web access and retrieval lower that rate without removing it, and systems that act rather than answer add failure modes of their own.

Energy and water cost

Course use runs to no more than about a thousand text prompt exchanges across the semester. On the first per-prompt figures Google has released, a median Gemini text prompt costs 0.24 watt-hours of electricity, 0.26 milliliters of water, and 0.03 grams of CO2, so the required use for this course comes to roughly a quarter of a kilowatt-hour, a cup of water, and 30 grams of carbon. Week 12 works on that figure at length.

Generating images and video is a different order of magnitude. Luccioni, Jernite, and Strubell measured 88 models across ten tasks and found image generation at about 2.9 watt-hours per image, a dozen text prompts each, and one five-second generated video drew about 944 watt-hours, close to four times this course’s entire semester of text use.

That matters for the Digital Storytelling Project, the one assignment where you might generate images or video. You may. Label them inside the piece, disclose them in the reflection, and the rule against synthetic depiction of real, identifiable people still applies. Cost is one consideration when you choose between generating a shot and filming one.

Tool choice and inputs

Partner and client information needs the right tool as well as care. U-M holds contracts covering U-M GPT and the university’s Google AI services, which is why they are approved for institutional data up to Moderate sensitivity and why what you type there is not used to train the models. Tools without a U-M agreement, including a personal ChatGPT or Claude account, are approved only for data classified as Low, and a consumer chatbot stores what you type, so an interview note or a client name entered there is no longer under your control.

In practice, anything touching an organization’s documents, internal materials, or the people it serves goes into U-M GPT or nowhere, and nothing above Moderate sensitivity, including protected health information, goes into any of them. De-identify before anything is entered, and if you cannot de-identify a document, do not run a tool on it. If you are unsure where something falls, check AI and U-M Data or ask me before you paste. The NASW Code of Ethics already requires this, both keeping client information confidential (1.07) and remaining competent with the technology you use (1.04), and Week 7 covers what these services do with the data they collect. What you build for a partner belongs to the partner; if you want it in your portfolio, we agree that at the intake conversation, while the terms are still open.

Reading load

Each week assigns three to five items, counting the course resource pages among them, and every one of them is linked from that week’s page in the course schedule. Videos and interactive pieces are treated as lighter items. Six books come up across the semester: Eubanks’ Automating Inequality, Benjamin’s Race After Technology, Noble’s Algorithms of Oppression, D’Ignazio and Klein’s Data Feminism, Costanza-Chock’s Design Justice, and Crawford’s Atlas of AI. The course assigns excerpts from them, and Data Feminism and Design Justice are fully open access online, so none of the six needs to be purchased.

Paywalled readings

About one reading in five is behind a publisher’s paywall, mostly the journal articles. The links on this site point at the publisher, and the library gives you two ways to open one of them from off campus.

• Put https://proxy.lib.umich.edu/login?url= in front of the address. A DOI link becomes https://proxy.lib.umich.edu/login?url=https://doi.org/10.1145/3442188.3445922.
• Or add the library’s Check for Library Access bookmarklet once and press it on any page that asks for payment.

If a reading is still unavailable after both, tell me and I will send a copy, since no week’s argument depends on a reading you cannot open.

Community organizations and guests

The course asks one hour of one organization. Each pair interviews one person who knows how the work runs there, once, for the Strategy Pitch, and the Week 9 lab builds the protocol, so that the hour returns something to the organization as well. Bring your own organization or ask me to recommend one, and name it by Week 6. Students on Track C of the Digital Story may also work with an organization or with footage from the Ginsberg Center archive, requested through me.

Two guest sessions run about an hour each. Lauri Goldkind (Fordham), editor of the Journal of Technology in Human Services, joins Week 6 on AI adoption inside human-services organizations, and Jes Kane, who worked on the Trevor Project’s crisis simulator from Google.org’s side, joins Week 8 on how product and funding decisions look from inside a technology company, and a response to each of the two counts toward participation.

Laptops

Bring one; half of every session is a lab. If that is a financial problem, talk to me or to SSW IT before the term starts, so you are not working around it all semester. During seminar, close whatever is not the reading or your notes. I will not police screens, but if you are plainly somewhere else during a discussion I will ask you about it.

Email

I answer within two business days. I do not read email overnight or on weekends, and you are not expected to either. If you write me the night before a deadline, assume I will miss it; spend a flex day instead.

Classroom conduct

This course covers deportation, child removal, benefits surveillance, and risk assessment, and some of you have lived through those systems. You are never required to bring your own experience into the room, and you are not the class expert on any group you belong to. I flag heavy material on the week page ahead of time; the material remains on the syllabus either way.

Project teams sometimes go badly, which is a professional problem and not a personal failure. If a partner or teammate is not doing their share, tell me by the midpoint of the project, while I still have room to fix it.

Anonymous feedback

There is an anonymous feedback form open the entire semester, asking three questions: what to continue, what to start, what to stop. Use it whenever something is working or not working. You do not have to wait for a survey window, you do not have to complete all three, and nothing you write is attached to your name. I read it as it comes in and tell the class what I am changing when I change it.

Recording

Ask me before you record a session. People here think out loud about deportation, child removal, and systems they have been through, and that depends on not wondering where a clip will be seen.

Permission is routine. If you have an accommodation through Services for Students with Disabilities that covers recording, send the documentation and you are set. If you need to record for any other reason, email me and we will sort it out. Retain the recording for your own use and delete it at the end of the term.

Course materials, including slides, packets, and session recordings, may not be reproduced or redistributed without my written consent.

Diversity and inclusion

Much of science is subjective and was historically built on a small subset of privileged voices, and technology and AI in particular have histories of excluding, misrepresenting, or erasing marginalized communities. There may be both overt and covert biases in the material because of the lens through which it was written. Integrating a diverse set of experiences matters for understanding how technology can serve, or harm, social justice.

I treat diversity, equity, and justice as core elements of the course. We examine how classification systems, data collection, algorithmic decisions, and design choices reflect and reinforce social hierarchies around race, gender, class, sexuality, disability, immigration status, language, and other identities. If you have suggestions for improving the course materials, tell me in person or by email, or use the anonymous feedback form. If something said in class made you feel uncomfortable, I welcome the chance to discuss it. I am still learning about perspectives and identities that are not my own, and that is a lifelong process. As a participant in discussion, you are also asked to honor the diverse perspectives and backgrounds of the class. If you would rather raise something with someone outside the course, the Office of Student Services in 1748 SSW is the school’s routing desk.

Names and pronouns

You will be called by the name and pronouns you use. If yours differ from the roster, tell me before the second session. You can set pronouns in Wolverine Access under Campus Personal Information.

Accommodations

Disability is part of diversity, and making this course accessible is my job. If you encounter a disability-related barrier, contact Services for Students with Disabilities at (734) 763-3000 or ssdoffice@umich.edu, and talk to me early so accommodations work from the start. The SSW Office of Student Services (ssw.oss@umich.edu) can also help.

School of Social Work resources

Start here. The Office of Student Services in 1748 SSW, ssw.oss@umich.edu or (734) 936-0961, is the school’s routing desk. If you do not know which office you need, that is the one to email; they also administer the SSW Student Emergency Fund for food, transportation, and housing needs.

Counseling. The School has its own embedded CAPS therapist who sees SSW students only. It offers what central CAPS offers, with a shorter intake.

Writing. SSW Writing Help runs individual appointments with the school’s Writing and Study Skills Coordinator, and it covers presentations and eportfolios as well as papers.

Careers. SSW Career Services handles resumes, licensure questions, and portfolio development. If the work you do for a community partner here is going into your portfolio, they are the ones to ask how to present it.

Technology. SSW IT Services supports SSW students directly, and the student technology page is the fastest index. Virtual Sites gives you SPSS, Stata, and MPlus at no cost without installing anything, and Study Spaces reserves a machine or a quiet room. For anything broken, the ITS Service Center is 734-764-HELP.

Cameras and audio kits. North Campus Technology Checkout, in Groundworks on the first floor of the Duderstadt Center, lends video and still camera kits, microphones, audio recording kits, lights, and gimbals to any student at no cost, weekdays 9 to 6. Reserve ahead of the Week 11 production lab; everyone in every media course wants the same kit the same week.

Studios. The School has a podcast studio in SSWB B696A set up for audio and video, open 9 to 9 Sunday through Friday and 9 to 4 Saturday, with a one-time orientation before your first booking. The SSW video studio in SSWB 2686 is built for single-presenter lecture capture and is primarily a faculty resource, though other setups such as a two-person interview are considered case by case; email SSW IT to ask. Across campus, the Shapiro Design Lab video production room in Shapiro Library 1074 is open to any student who finishes its short Canvas training, and it has a tripod rigged for phones and cameras, wired and wireless mics, and a Mac Studio running Adobe Creative Suite and Final Cut. You pick up the room key at the lobby desk. If you want someone to help with lighting or sound on the day, email shapirodesignlab@umich.edu at least 72 hours ahead. The Stamps studios are for Stamps students, so plan on the Duderstadt and Shapiro spaces instead.

Community partnerships. The Ginsberg Center advises on equitable community partnerships and is a useful outside read if your project raises something you would rather not take to your partner first.

Religious observance, military duty, and everything else

Reasonable arrangements are made for religious obligations that conflict with course requirements; tell me in advance. Students called to active duty should contact me and the school immediately. For safety procedures, mental health and wellbeing resources, writing support, teaching evaluations, and academic integrity, see the SSW standard policies page. Plagiarism rules apply to AI-generated content exactly as they do to any other unattributed source.