The city wants a culture that sustains responsible AI use. Which cultural goal best supports that?

For a culture that sustains responsible AI use across city departments, the cultural goal that best supports that outcome is rewarding responsible experimentation and documented learning—not only flashy launches. What leaders celebrate becomes what staff repeat. Recognizing careful trials, measured results, and write-ups of what failed safely encourages the behaviors municipal AI needs: curiosity bounded by policy, evidence, and resident protection. Rewarding only silent failure so problems stay invisible fails by incentivizing concealment and guaranteeing larger incidents later. Punishing any experiment that does not succeed on the first try fails by freezing innovation into either paralysis or covert shadow tools outside governance. Celebrating launches exclusively and ignoring post-launch learning fails because launch theater without learning recreates pilot-and-abandon cycles that leave orphaned systems connected to city data. Exam caveat: culture questions often hide in HR-sounding language; look for incentives aligned to responsible learning rather than announcement volume. Operational check: add recognition criteria for documented experiments and post-launch learning reviews, include them in performance conversations, and stop treating go-live announcements as the only success story in city communications.