Henkel, MoritzMoritzHenkelKull, PascalPascalKullMusawi, AnuschAnuschMusawiRemmen, Judith vanJudith vanRemmenFörster, FelixFelixFörsterGötz, StefanStefanGötzBursac, NikolaNikolaBursacWartzack, SandroSandroWartzackRitzer, KatharinaKatharinaRitzer2026-10-272026-10-272026-1037. DfX-Symposium 2026https://hdl.handle.net/11420/65536Collaborative design teams in team flow exhibit high communicative coherence but may suppress critical reflection, yielding poor decision quality despite apparent effectiveness — a phenomenon the authors term "Deceptive Fluency." This exploratory study combines AI-based team flow analysis (Teamflow-Score 6.83/10) with expert-rated decision quality assessment (S1=2.48, S2=2.04, Fleiss' κ=0.57, nominal) in a real GSD live-lab session. Results from this single case study provide first exploratory indications of a co-occurrence of high team flow and low decision quality in 7 of 9 decisions. Ten requirements for a data-driven, flow-preserving assistance system are derived across three dimensions: recognition, intervention design, and technical integrationdeteam flowdesign decision-makingcollaborative designAI assistance systemdeceptive fluencyTechnology::600: TechnologyConference Paper10.35199/dfx2026.24