Ethics and Security

AITECH · 45 questions

• A 311 resident-services desk worries a benefits chatbot answers differently when residents write in neighborhood dialect. Which responsible-AI action best addresses that risk?
• A transit authority launches an eligibility assistant for reduced fares and wants riders to understand how it works. Which responsible-AI practice best supports transparency?
• A water utility begins using model scores to help prioritize housing waitlist outreach. What accountability practice should leadership establish first?
• An emergency-management office co-sponsors a youth library chatbot and must keep it from giving self-harm advice. Which responsible-AI control best matches that safety need?
• A housing authority notices its AI help examples and evaluation sets are almost all English. What bias-mitigation step best improves language-access equity?
• A public library system uses AI to flag possible code violations in reserved meeting spaces. Which accountability practice keeps residents able to contest those flags?
• A parks and recreation department deploys a flood-risk summarizer for planners. How should it practice transparency about known limits?
• Municipal courts staff want to know whether a permit-timing predictor errs more in some ZIP codes than others. Which fairness evaluation practice should they use?
• A fire department IT team is asked to let an AI deny resident assistance benefits with no human review. What responsible deployment boundary should they set?
• A police records bureau is drafting AI use principles for internal summarization tools. How should those principles connect to existing city policy?
• A city open-data portal team hears its chatbot sometimes stereotypes neighborhoods. What bias-discovery practice should they run?
• School-district IT notices staff justifying poor decisions with “the model said so.” What accountability message should training reinforce?
• An airport authority finds staff pasting case notes that include Social Security numbers into a public AI chat tool. Which privacy control best stops that exposure?
• A port authority must summarize meeting transcripts with a cloud model but wants to limit personal data leaving the agency. Which preprocessing step best protects privacy?
• A county assessor’s office wants technical controls when staff paste medical record numbers into AI prompts. Which security control best fits?
• A public works yard needs Q&A over juvenile-related records and cannot accept consumer-cloud training on that text. Which hosting choice best serves privacy?
• An election office is selecting a generative AI vendor for internal drafting. Which contractual privacy review should it prioritize?
• An animal-control division stores internal policy drafts as embeddings in a vector database used by staff AI search. How should it protect that data at rest?
• A solid-waste utility shares one long-lived AI API key across teams, and a contractor just left. What access-security practice should it adopt?
• A stormwater bureau finds unapproved browser extensions sending clerk-screen text to outside AI services. What control best prevents that uncontrolled leakage?
• A building inspection division builds a simple permit-status FAQ bot and must decide which case fields to send to the model. What privacy practice should guide that design?
• Planning commission staff maintain both public open-data corpora and confidential case corpora for AI retrieval. What control best protects sensitive civic data?
• A city CIO office finds staff pasting resident data into consumer chatbots that are not city-approved. Which organizational security control addresses that risk?
• A municipal broadband team operates an AI assistant that can retrieve resident account PII for authorized clerks. What control best supports auditability of that access?
• Public hospital IT worries that a forged AI-generated voice could spoof emergency alert messages to staff. Which AI-specific risk is that primarily describing?
• A veterans services office notes that attackers can hide instructions inside user text to override a benefits chatbot’s rules. How should they distinguish that threat?
• A community development agency finds chatbots telling residents that emergency shelters are closed when they are open. What AI-specific risk does that illustrate?
• Code enforcement considers letting public form uploads enter a continuous training set for a citation-triage model. Which AI-specific risk should they assess first?
• A parking authority worries that crafted adversarial prompts could pull internal enforcement-policy text out of an assistant that has that corpus in context. What threat is that?
• Convention-center IT publishes AI-generated promotional images for public events. What practice best reduces deceptive reuse of those visuals?
• A zoo and parks wildlife unit notices staff trusting fluent AI summaries of wildlife ordinances that later prove incorrect. What risk should leadership highlight?
• Museum district IT is planning the annual cybersecurity tabletop exercise. How should they treat AI-specific risks?
• A county notices mass AI-generated fake comments flooding a zoning hearing portal. Which AI-specific risk does that primarily represent?
• Social-services intake security training must separate traditional malware risk from generative content risk. Which statement captures that distinction?
• Child-protective services IT monitors public kiosks that run a filtered AI helper. What AI security risk should they watch for?
• A homeless-services continuum briefs leadership on deepfake videos falsely portraying officials canceling winter shelter funding. Why is that briefing relevant to AI ethics and security?
• Public defender IT is drafting rules for how attorneys may use generative AI with case materials. What does an effective AI governance policy typically define?
• A city attorney office wants AI pilots tracked like other enterprise initiatives. What risk-management practice supports that goal?
• A budget office is preparing to launch a resident benefits chatbot. What compliance step should governance require before go-live?
• Fleet management wants departments to stop spinning up shadow GenAI pilots without review. What governance mechanism helps?
• Facilities maintenance proposes an AI assistant that may touch employee health-accommodation records. What governance gate is most appropriate?
• A GIS mapping unit’s AI tool can suggest parcel designations that may affect property rights. What governance control should be defined?
• The city CISO team wants AI problems such as data leakage or biased outcomes handled consistently. What governance practice supports that?
• A records retention office is evaluating vendors for a generative AI records assistant. How should procurement support AI governance?
• An ADA compliance office owns a production accessibility chatbot whose model and prompts change over time. What governance practice should they set?