Key question New
AI-processed · Author-approved
How can teams integrate generative AI into creative problem-solving without losing human agency and control?
Loads Spotify's player; Spotify may set cookies. This device remembers your choice.
Key findings
- Generative AI accelerated data synthesis, user need clustering, and idea generation, but faster speed and greater volume of ideas did not reliably lead to higher-quality concepts compared to control groups.
- Less structured AI use was associated with cognitive fixation, reduced critical deliberation, and weaker feelings of team ownership over the final creative solution.
- Generative AI was most productive when assigned to bounded, analytical tasks while human teams retained responsibility for problem framing, idea selection, and final concept decisions.
- AID Thinking structured human-AI creative collaboration through explicit governance mechanisms, including task and role allocation, prompt scaffolding, and human agency checkpoints.
- Governance routines were enacted unevenly and benefited from active facilitation, prompt guidance, templates, and structured documentation.
What it means for you
CIO / IT Executive
Establish governance structures that help teams decide where GenAI can add value in creative work while keeping human responsibility for framing, evaluation, and consequential decisions explicit.
IT Manager
Provide prompt guidance, structured documentation, and explicit human checkpoints to help teams critically evaluate AI contributions and retain ownership of final concept decisions.
Business Strategist
Design creative workflows with explicit human agency checkpoints at which teams critically evaluate, adapt, or reject AI-generated contributions before proceeding.
Researcher
Investigate how structured work-design interventions like AID Thinking impact the trade-off between AI-driven content acceleration and human critical deliberation in collaborative creative teams.
Policymaker
Develop guidance that keeps human accountability explicit when generative AI contributes to collaborative creative work, particularly for consequential evaluation and design decisions.
Summary
This study develops and evaluates AID Thinking, a work-design intervention format designed to integrate generative AI into collaborative innovation activities. Following a four-cycle design science research methodology, the researchers synthesized Design Thinking challenges with AI capabilities and tested the format across three workshop cycles involving 44 participants. Data collection combined team design outputs, prompt logs, pre- and post-surveys, and interviews to examine how teams structure human-AI collaboration.
Problem
While generative artificial intelligence can accelerate content creation and information analysis, introducing it into collaborative creative work creates socio-technical risks. Teams can face issues such as overreliance, design fixation, reduced critical discussion, and ambiguous accountability for final design choices. Without structured collaboration design, simply giving teams access to AI tools may weaken human agency and does not necessarily improve creative output quality.
Transcript
Expert: Hello, Anna. It is great to be here.
Host: Today we are discussing a study titled "From Design Thinking to AID Thinking: A Work-Design Intervention Format for Generative AI-Augmented Creative Teams," authored by Stephan Leible, Dejan Simic, Furkan Dursun, Florian Straetmanns, and Constantin von Brackel-Schmidt, published in the proceedings of the International Conference on Information Systems, 2026. Alex, could you share what this study is about?
Expert: The authors developed and evaluated a work-design intervention format called AID Thinking, designed to integrate generative AI into collaborative innovation activities.
Host: Many organizations introduce generative AI tools into creative teams expecting immediate productivity gains. What challenges does the study examine?
Expert: The study highlights risks such as overreliance on AI outputs, cognitive fixation, reduced critical discussion, and unclear accountability. While generative tools speed up data synthesis and brainstorming, faster generation and higher volume of ideas do not reliably result in higher-quality concepts. Without intentional work design, introducing these tools can weaken team agency.
Host: How did the researchers conduct this study?
Expert: The research followed a four-cycle design science research methodology. The authors evaluated the format across three workshop cycles with 44 participants, analyzing design outputs, prompt logs, pre- and post-surveys, and participant interviews to track how human-AI collaboration actually unfolded.
Host: What were the key findings regarding team dynamics and creative output?
Expert: When AI use was insufficiently structured, some teams relied heavily on AI suggestions with limited critical deliberation, while participants also reported generic ideas, fixation, and weaker ownership.
Host: That is a significant issue for organizations seeking original outcomes. Did the study identify conditions where collaboration was more effective?
Expert: Yes. In contrast, more productive collaboration patterns emerged when teams used AI for bounded analytical or structuring tasks while humans retained responsibility for problem framing, idea selection, and final design decisions. When generative AI was restricted to tasks like clustering needs or synthesizing data, human teams retained ownership of the core creative directions.
Host: That brings us directly to AID Thinking. How does this intervention format structure that collaboration?
Expert: AID Thinking provides explicit governance mechanisms. It organizes task and role allocation, provides prompt scaffolding, and incorporates human agency checkpoints where teams must pause to reflect, critique, and decide on their direction before moving forward.
Host: Can organizations expect teams to adopt these governance practices on their own once given access to generative AI?
Expert: The researchers also found that these governance practices were enacted unevenly across teams. Their implementation therefore benefits from structured templates, prompt guidance, documentation routines, and active facilitation.
Host: That provides a clear takeaway for practitioners: successful integration is not just about tool access, but about designing structured processes that preserve human accountability and deliberate decision-making. Thank you, Alex, for walking us through this study.
Expert: You are very welcome, Anna.
Host: And thank you to our listeners for tuning into A.I.S. Insights — powered by Living Knowledge. Join us next time as we continue examining evidence-based insights in business and technology.
Generative AI, Design Thinking, Human-AI Collaboration, Creative Knowledge Work, Creative Teamwork, Work Design