Synthetic Image, Real Harm: Responding When the Policy Has a Gap
At a suburban middle school, a counselor receives a screenshot from a group chat of 14 seventh graders. The image is an AI-generated picture of a named classmate placed in an embarrassing setting. Three students appear to have created and shared it, and at least nine others forwarded it. Two of the three creators used personal phones on home Wi-Fi. The campus acceptable-use policy covers 'inappropriate use of district devices' but says nothing about synthetic media or off-campus distribution. The principal asks the technology specialist for a recommendation before the end of the day. State law requires notifying a target's parents when a student is the victim of bullying, but the district's bullying definition has not been revised in five years.
1. Lay out the response sequence for the next several hours. Justify why the steps go in that order.
Read the worked answer
Order matters because evidence preservation and student protection pull in different directions. First, preserve: have the counselor save the screenshot with its metadata and note sender and time, and stop students from deleting the chat before that record exists. Second, protect: check on the targeted student, notify that student's parents under the state victim-notification requirement, and offer counseling; do not interview the target in front of peers or ask them to re-describe the image. Third, investigate narrowly: interview the three creators separately with an administrator present, asking what was made, with which tool, and who received it; document rather than debate whether the image is 'real.' Fourth, apply the code of conduct to conduct with a school effect: creating and forwarding a degrading sexualized image of a classmate creates a hostile environment regardless of where the phone was. Do not promise specific consequences before the AUP gap is addressed, and route the discipline question through district legal.
2. One creator's parent argues: 'It happened at home, on my phone, so the school has no jurisdiction.' How should the specialist respond?
Read the worked answer
Concede what is true and hold what is not. The parent is correct that a district device policy does not reach a home device, and pretending otherwise destroys credibility. But three things remain within school authority: the image was distributed into a student group chat and its effects surfaced at school, so harassment and conduct rules reach the student's school-day behavior, including forwarding, screen-sharing, and discussing it; the target's ability to attend class safely is a school obligation under bullying and harassment rules, and possibly Title IX if the image is sexualized; and the state's victim-notification duty is triggered by the conduct, not by the location of the phone. Then name the gap honestly: the acceptable-use policy's silence on synthetic media is a district defect, not the parent's fault, and it will be fixed. Offer something specific and restorative (verified deletion, an apology plan, a media-literacy consequence) while being clear that off-campus does not mean consequence-free.
3. The school's digital citizenship program is a single October assembly. Propose a redesign and explain how you would know it worked.
Read the worked answer
One assembly fails on two counts: it is massed practice with no retrieval, and it teaches rules rather than the judgment students needed at the moment they hit 'forward.' Redesign as three short, spaced cycles of about 20 minutes, one per month, using the structure of this incident without the individuals: who is harmed, who is responsible, what you do at the forward button, and what the law actually says about synthetic images of minors. Include the technical layer, because students who understand the mechanism make better calls: how generation tools work, why 'it's obviously fake' is not a defense, and how screenshots and metadata persist. Add a performance task in which students draft and defend a one-page classroom norm for AI-generated images. Measure with a pre/post scenario set rather than a satisfaction survey, and track the outcome that actually matters: whether targets and bystanders tell an adult.