The Hidden Cost of AI
While Generative AI brings massive productivity boosts, it also introduces a hidden risk to our organizations: “deskilling”. When we rely too much on AI systems to do the thinking for us, we risk eroding our own decision-making capabilities, domain expertise, and independent reasoning.
Currently, most AI training focuses on “prompt engineering”—trying to craft the perfect input to get the right output in a single try. However, real-world work requires ongoing iteration and human judgment. If we blindly accept fluent AI outputs without critical evaluation, we suffer from “authorship drift”. This means we gradually outsource our goal-setting, evaluative judgment, and responsibility to the machine. We need a new way to work with AI that prevents this loss of human agency.
Frame. Explore. Refine. Commit.
This structure trains users and organizations to treat AI as a co-creative assistant rather than an oracle. Pride is placed in reflective evaluation and strategic direction, not autonomous production.
Figure: The FERC Framework and FERC application principles.
FERC fundamentally shifts the focus from optimizing an AI’s artifact to governing the process of interaction. By structurally safeguarding human intent, judgment, and responsibility, the framework ensures that humans remain the true authors of the work, even as the AI handles the heavy lifting of generating content.
The FERC-Bot
Because conceptual frameworks alone rarely change everyday habits, we developed the FERC-Bot. It embeds the four FERC stages directly into a conversational interface to act as a training system. As you work, you explicitly select your current stage, and the bot provides real-time feedback. It actively scores your interaction—measuring, for example, whether you actually requested multiple alternatives in the Explore stage or conducted a rigorous comparative evaluation in the Refine stage.
Image: Visualization of the FERC-bot
Research
The FERC framework is deeply grounded in cognitive science and creativity research. Research shows that true creativity and problem-solving are not one-shot generation events, but an iterative interplay between generating ideas and evaluating them. FERC operationalizes this “metacognitive control” for hybrid intelligence systems.
The framework reframes classic problem-solving models, like the Double Diamond design model, specifically for AI collaboration. It also introduces a vital distinction between ownership (who contractually owns the final output) and authorship (who exercises accountable agency over the creative process). Tested across executive training contexts, our research demonstrates that making collaboration structures visible and measurable is key to preventing AI-induced deskilling and preserving human expertise.
Ensuring Human Authorship in the Age of Generative AI
FERC as a way for Hybrid Intelligence
In the era of generative AI, the default mode of interaction often slips into passive delegation—we type a prompt, the algorithm produces an output, and human authorship gradually fades. While this approach may offer short-term efficiency, it fundamentally undermines long-term organizational value and human agency.
By structurally separating human judgment from machine generation, the FERC framework prevents organizations from falling into the “efficiency trap” of pure automation. It transforms AI from a tool that substitutes human labor into a partner that elevates human potential. FERC ensures that as our technology becomes more autonomous, our workflows remain human.
2000+ has already been trained in using the FERC Framework
Including organizations like …
Academic Contributors
Jacob F. Sherson, Janet Rafner, Roni Reiter-Palmon, Izabela Lebuda, Matthias Söllner, Yoed Nissan Kenett, Blerim Emruli, Manuel Rindle, Selina Weiss, Benjamin Goecke, Florent Vinchon, Yaoli Mao, Dominik Dellermann, Seyedahmad Rahimim, Janet H. Marler, Andy Nguyen, Jens Christian Bjerring, Steve Dipaola, Frederik Brosbøl Kjeldsen