Swiss AI Academy Launches Framework to Keep Humans in Charge as AI Scales

Swiss AI Academy has launched the Bionic Context Protocol (BCP), a 14-principle framework designed to prevent the erosion of human judgment and capability as artificial intelligence spreads across societies worldwide. 

The announcement was made during an independent auxiliary unDAVOS event alongside the World Economic Forum Annual Meeting.

"How you use AI matters as much as whether you use AI," Shaje Ganny, (PICTURED) Co-Founder of Swiss AI Academy, told That's Business. 

"When people passively accept AI outputs, capabilities degrade. When AI is designed to keep humans thinking and challenging, capabilities strengthen."

The problem: adoption outpacing safeguards

AI adoption is moving faster than governance and human capability safeguards. A 2025 MIT Media Lab study found that people who relied on AI writing tools showed weaker brain connectivity and struggled to recall their own work, a phenomenon researchers termed "cognitive debt." Current responses remain siloed: researchers study the problem, ethicists debate principles, organizations develop internal policies. BCP unifies this work into a coherent protocol that moves from discussion to implementation

The evidence: four decades of automation research

BCP is built on research into automation bias, skill decay, and human-machine interaction from safety-critical industries including aviation and healthcare. Studies from these fields show that when humans become passive observers of automated systems, their ability to intervene during failures declines.

The framework: three levels of protection

The protocol operates at three levels: individual, protecting personal agency and independent thinking; organisational, ensuring human needs are not subordinated to efficiency metrics; and societal, preserving the capacity of communities to shape their collective future.

The distinction: evolution versus erosion

BCP distinguishes between capability evolution, where societies intentionally choose which skills to develop or retire, and capability erosion, where skills disappear as an untracked side effect of systems optimized for speed or cost.

The call: global recruitment for five workstreams

The framework is released as version 0.6, a consultation draft intended to be completed through public contribution. Swiss AI Academy is recruiting workstream leaders and contributors across five areas: governance architecture, evidence synthesis, implementation tools, measurement systems, and sector-specific applications.

"A small group cannot carry this alone," Ganny said. "We need researchers, practitioners, educators, and policymakers who understand what is at stake."

The full framework and contributor registration are available at bcporg.info.